Fort Fun Indiana
Diamond Member
- Mar 10, 2017
- 120,332
- 124,744
- 3,645
That's a great amp. I have the 90w version.
I'm tired of anemic guitar amps!
- Thread starter scruffy
- Start date
- Thread starter
- #122
You are right, solid state is cleaner and less lossy (plus it doesn't need an extra 5v winding on the PT).
Some of the blues players like the "sag" from tube rectifiers, done right it's kind of like a poor man's compression. I have no use for that, I like my amps attacky and dynamic.
- Thread starter
- #123
Well, I just found out the Dual Rectifier uses Soldano's magic circuit, they call it a "cold clipper". Apparently the concept started with the JCM 800 #2203, but Marshall used a 10k cathode resistor which means you only get the effect near full gain.
Haven't tried the Dual Rectifier myself, never been much of a Mesa fan, people seem to like them though.
- Thread starter
- #124
Just punched the octal holes for the capacitors, tomorrow I should be almost fully drilled. The transformer grommets can go in and the can mounts for the caps.
Then it's just stripping the rest of the paint off the back panel and I can start building. I'm hoping the tube sockets still work after all these years, the small ones are riveted in and I don't relish the thought of drilling them out. Maybe need a little DeOxit or something, we'll see.
Then it's just stripping the rest of the paint off the back panel and I can start building. I'm hoping the tube sockets still work after all these years, the small ones are riveted in and I don't relish the thought of drilling them out. Maybe need a little DeOxit or something, we'll see.
Point made, but it is possible to model that "sag" with "current limiting" circuits. One could even make it adjustable.. just thinking out loud. Good insights!
Last edited:
It's been too long to remember the nomenclatures; we have several types and power ranges. But they were no amplifier tubes. They actually generated or "oscillated" the RF needed internally, Like and oscillator and amp build in one tube. Some were clear glass, like the one pictured, and others were ceramic. It was an art form to control, direct and tune (dial one in.) Especially after installing a new tube.
- Thread starter
- #127
Eimac 4-1000A VINTAGE TRANSMITTING ELECTRON TUBE EIMAC Read | eBay
Was told told that the tube has life but is weak. Sold as is. Everything you see in the pictures is what is included in this lot. We of course will get them out as fast as we can. We will help resolve whatever issue you are having.
www.ebay.com
I built one of these when I was a kid, it's a kilowatt linear amplifier for amateur radio
Looks kinda like this, except this tube is its little brother 3-400Z:
- Thread starter
- #128
Mesa Dual Rectifier circuit quite obviously stolen directly from Soldano. They're brazen about it, same part values and everything.
You'll notice a few things:
1. The 1 uF cathode bypass caps give you an 800 Hz rise very similar to Marshall's .68 uF in the lead channel. In my circuit, I switch the value of this cap through a range that includes 1 uF for lead, 4.7 uF for country, 25 uF for Fender and 220 uF for bass.
2. The primitive tone shaping network between stage 1 and stage 2. When LDR4 is activated the composite values are identical to Soldano, 470k and .002 uF. Marshall uses either a slightly bitier cap in the lead channel or no cap at all in the normal channel. In my circuit, I replace the cap with an Orange-style switchable coupling cap, allowing very precise control of low frequencies.
3. The 39k "cold clipping" resistor in the third stage cathode. In my circuit I switch this 4.7k-10k-22k-39k, and there is a dramatic change in level at each setting, and in my circuit there is a reverb after the cathode follower which means you have to compensate for these changes in gain. The Mesa circuit will deliver approximately 22 volts at the output of the cathode follower, which is way too much for a reverb. So in my circuit there is a Dwell control like in the Fender 6G15, you can drive the reverb as hard or as quietly as you want. Also Mesa's effects loop output is ridiculously high, you'll notice that Soldano drops this by a factor of 100 (I use 50, same principle).
4. Mesa's red channel "presence" control is kloodgy, it doesn't give you the same hair that a real presence control does. The only way to get that is to use the NFB. For some reason Mesa only does this in the orange channel. In my circuit I switch the presence cap and also use an RC network in the feedback loop to provide precise control over both treble and bass frequencies.
In my dial-an-amp circuit, you get initial control over the bass frequencies in the first stage, via the cathode cap and the coupling cap. That signal then goes into a Fender-style tone stack that adds shimmer to the highs. Then that goes into the crunch stages, followed by a Marshall style tone stack. The idea is you want to control very carefully what goes "into" your crunch, and then shape the resulting tone afterwards. You can get different characters of thump by adjusting the bass rolloff going into the crunch. Too much bass makes it sound buzzy, whereas not enough makes it bitey. Bass thump is somewhere in the 100 Hz area for a guitar, and midrange thump could go as high as 500 Hz. You want rolloff so anything below the thump frequency gets attenuated (which means, stays clean and has lower volume).
KT-88's deliver an immensely powerful thump throughout this entire frequency range, but I keep high value coupling caps in the output chain so a bass guitar can still be clean (with no obvious "thump" peaking anywhere in the midrange). For bass guitar I switch out the cathode follower entirely, so the output stage just gets a loud clean signal that can be cold-clipped as desired. One of the presence options is tuned to the string noise frequency of roundwounds on a bass, so you can get as much of it or as little of it as you want.
I'll put my amp up against a Dual Rectifier any day, mine is infinitely superior with more tonal options and better dynamics. Makes a Marshall sound like a toy.
You'll notice a few things:
1. The 1 uF cathode bypass caps give you an 800 Hz rise very similar to Marshall's .68 uF in the lead channel. In my circuit, I switch the value of this cap through a range that includes 1 uF for lead, 4.7 uF for country, 25 uF for Fender and 220 uF for bass.
2. The primitive tone shaping network between stage 1 and stage 2. When LDR4 is activated the composite values are identical to Soldano, 470k and .002 uF. Marshall uses either a slightly bitier cap in the lead channel or no cap at all in the normal channel. In my circuit, I replace the cap with an Orange-style switchable coupling cap, allowing very precise control of low frequencies.
3. The 39k "cold clipping" resistor in the third stage cathode. In my circuit I switch this 4.7k-10k-22k-39k, and there is a dramatic change in level at each setting, and in my circuit there is a reverb after the cathode follower which means you have to compensate for these changes in gain. The Mesa circuit will deliver approximately 22 volts at the output of the cathode follower, which is way too much for a reverb. So in my circuit there is a Dwell control like in the Fender 6G15, you can drive the reverb as hard or as quietly as you want. Also Mesa's effects loop output is ridiculously high, you'll notice that Soldano drops this by a factor of 100 (I use 50, same principle).
4. Mesa's red channel "presence" control is kloodgy, it doesn't give you the same hair that a real presence control does. The only way to get that is to use the NFB. For some reason Mesa only does this in the orange channel. In my circuit I switch the presence cap and also use an RC network in the feedback loop to provide precise control over both treble and bass frequencies.
In my dial-an-amp circuit, you get initial control over the bass frequencies in the first stage, via the cathode cap and the coupling cap. That signal then goes into a Fender-style tone stack that adds shimmer to the highs. Then that goes into the crunch stages, followed by a Marshall style tone stack. The idea is you want to control very carefully what goes "into" your crunch, and then shape the resulting tone afterwards. You can get different characters of thump by adjusting the bass rolloff going into the crunch. Too much bass makes it sound buzzy, whereas not enough makes it bitey. Bass thump is somewhere in the 100 Hz area for a guitar, and midrange thump could go as high as 500 Hz. You want rolloff so anything below the thump frequency gets attenuated (which means, stays clean and has lower volume).
KT-88's deliver an immensely powerful thump throughout this entire frequency range, but I keep high value coupling caps in the output chain so a bass guitar can still be clean (with no obvious "thump" peaking anywhere in the midrange). For bass guitar I switch out the cathode follower entirely, so the output stage just gets a loud clean signal that can be cold-clipped as desired. One of the presence options is tuned to the string noise frequency of roundwounds on a bass, so you can get as much of it or as little of it as you want.
I'll put my amp up against a Dual Rectifier any day, mine is infinitely superior with more tonal options and better dynamics. Makes a Marshall sound like a toy.
- Thread starter
- #129
Well, I managed to get the lettering off the front panel. Now it's just matte black. The black part is some kind of weirdo powdercoating, it's not anodized but it sure is persistent. It's resistant to 220 sandpaper, it does come off but ever so slowly. I'd like to get it down to the bare metal, haven't figured out where to go from there yet, but one step at a time... 
The chassis is all drilled out, the only thing that remains is to grind out the hole for the IEC socket (that'll be tomorrow). There'll be a few more holes as time goes by, nothing major though. The hard part is done, that was the cap cans, 1-1/8" punch holes. And I decided to put an extra tube (socket) in, that was 7/8".
So now... there's 560 volts on the HT, and the downstream caps are only rated at 500 volts. So we have to make sure that the voltage at the top of the cap chain never exceeds 500. The first resistor is supposed to drop the voltage from 560 to 450, but that only works if the tubes are working. If we pull the preamp tubes, that voltage wants to go back to 560, so we need a large resistor to ground that will double as a bleeder for the cap chain when the standby switch is out.
Natively with the tubes in, an 8.2k resistor takes us down to 450 volts, so what we can do is split that into 70v (so we get 490 which leaves a tiny safety margin), and an additional 40v. To do that we need 70/110 = 63.6% of 8.2k, which works out to 5218 ohms. The nearest available value is 5.6k. We can do a little math and calculate the current, by Ohm's law E=IR so with 160 volt drop at 8.2k that gives us .0195 amps, so round up to 20 mA. If zero current is being drawn downstream (worst case) we therefore need to drop 500 volts which means 9.75 watts. However if zero current is being drawn we can live with 1 mA as long as the voltage doesn't exceed 500. So 560 volts at 1 mA gives us 560k, which is the top of the range, and in this case we round down instead of up because we want to go to 500 volts "or less". Since we already have 5.6k our minimum value becomes 500/60 ° 5.6k or about 8.3 times the dropper value, which gives us a very convenient 47k. The current there is 10.6 mils, which is a little over 5 watts.
So we can very safely use two 27k 10w resistors in series, for a total of 54k. Adding in the 8.2k gives us 62k. At 560 volts the total current draw through the droppers is therefore 9 mA. With this configuration our baseline current becomes 30 mA instead of 20, but we're guaranteed never to exceed 500 volts at the top of the capacitor chain. Plus this gives us a 56k bleeder value in standby mode.
This is a total nit for the PT, it has no.impact whatsoever. At full current draw of 320 mils the PT still has 180 mils to spare. So basically we're chewing up 10 watts just to keep our capacitors safe. That's actually not too bad as far as tube amps go, no one cares about an extra ten watts.
The chassis is all drilled out, the only thing that remains is to grind out the hole for the IEC socket (that'll be tomorrow). There'll be a few more holes as time goes by, nothing major though. The hard part is done, that was the cap cans, 1-1/8" punch holes. And I decided to put an extra tube (socket) in, that was 7/8".
So now... there's 560 volts on the HT, and the downstream caps are only rated at 500 volts. So we have to make sure that the voltage at the top of the cap chain never exceeds 500. The first resistor is supposed to drop the voltage from 560 to 450, but that only works if the tubes are working. If we pull the preamp tubes, that voltage wants to go back to 560, so we need a large resistor to ground that will double as a bleeder for the cap chain when the standby switch is out.
Natively with the tubes in, an 8.2k resistor takes us down to 450 volts, so what we can do is split that into 70v (so we get 490 which leaves a tiny safety margin), and an additional 40v. To do that we need 70/110 = 63.6% of 8.2k, which works out to 5218 ohms. The nearest available value is 5.6k. We can do a little math and calculate the current, by Ohm's law E=IR so with 160 volt drop at 8.2k that gives us .0195 amps, so round up to 20 mA. If zero current is being drawn downstream (worst case) we therefore need to drop 500 volts which means 9.75 watts. However if zero current is being drawn we can live with 1 mA as long as the voltage doesn't exceed 500. So 560 volts at 1 mA gives us 560k, which is the top of the range, and in this case we round down instead of up because we want to go to 500 volts "or less". Since we already have 5.6k our minimum value becomes 500/60 ° 5.6k or about 8.3 times the dropper value, which gives us a very convenient 47k. The current there is 10.6 mils, which is a little over 5 watts.
So we can very safely use two 27k 10w resistors in series, for a total of 54k. Adding in the 8.2k gives us 62k. At 560 volts the total current draw through the droppers is therefore 9 mA. With this configuration our baseline current becomes 30 mA instead of 20, but we're guaranteed never to exceed 500 volts at the top of the capacitor chain. Plus this gives us a 56k bleeder value in standby mode.
This is a total nit for the PT, it has no.impact whatsoever. At full current draw of 320 mils the PT still has 180 mils to spare. So basically we're chewing up 10 watts just to keep our capacitors safe. That's actually not too bad as far as tube amps go, no one cares about an extra ten watts.
- Thread starter
- #130
So then, reviewing this calculation again, I propose to use a 5,6k 10 watt resistor feeding the capacitor chain from an initial input of 560 volts. The estimate said this should give use about 485 volts at the top of the chain, and to be extra safe I propose to use two 22k 5w resistors in series, as bleeders from the top of the chain.
Using Ohm's law, total current through this pathway is 11 mA. (560 volts at 50.6k total). The voltage at the top of the cap chain is therefore 560 - (5.6k * ,011 A), which is 498.4 volts worst case. The two 22k resistors will therefore need 500 * .011 = 5.5 watts total, so a rating of 5 watts each is more than sufficient.
Check. Now we just have to make sure we're still in range when the amp turns on, when the 11 mA draw becomes 31 mA. The danger is that the 5,6k resistor will now drop too much voltage. And sure enough, 5.6k at 31 mA is 174 volts, which leaves us 386 volts which is not enough. So we have to decrease the value of all the resistors, and chew up a few more watts to make sure we still have 450 volts at the top of the chain.
But you can see how the calculation goes. I'll run it through the linear solver tonight and get the correct values. Basically we want the smallest possible difference between standby mode and non-standby mode, with the two constraints that voltage at the top of the chain must always be less than 500 and greater than 450.
Using Ohm's law, total current through this pathway is 11 mA. (560 volts at 50.6k total). The voltage at the top of the cap chain is therefore 560 - (5.6k * ,011 A), which is 498.4 volts worst case. The two 22k resistors will therefore need 500 * .011 = 5.5 watts total, so a rating of 5 watts each is more than sufficient.
Check. Now we just have to make sure we're still in range when the amp turns on, when the 11 mA draw becomes 31 mA. The danger is that the 5,6k resistor will now drop too much voltage. And sure enough, 5.6k at 31 mA is 174 volts, which leaves us 386 volts which is not enough. So we have to decrease the value of all the resistors, and chew up a few more watts to make sure we still have 450 volts at the top of the chain.
But you can see how the calculation goes. I'll run it through the linear solver tonight and get the correct values. Basically we want the smallest possible difference between standby mode and non-standby mode, with the two constraints that voltage at the top of the chain must always be less than 500 and greater than 450.
- Thread starter
- #131
The linear solver says 2.2k and 18k. (I used Pulp in Python).
We can verify that. With no downstream draw, that means we'll be dropping 560 volts through 20.2k, which means we'll have 499 volts at the top of the capacitor chain. Check. Current draw here will be 27.7 mA.
At full downstream draw of 20 mA that gives us 47.7 mA total, and therefore the voltage drop through a 2.2k resistor is 104.94 volts, leaving 455 volts at the top of the chain. Check.
So then, wattage through 18k at 500 volts is 13.85 watts, so we can use a 10k and an 8.2k in series at 10 watts each. Problem solved.
Ain't that interesting, these values are almost identical to what Marshall uses in their 200-watter. They use a 1.8k dropper and a somewhat complicated arrangement of 3 sets of bleeders that works out to about 38k. So they're "slightly above" 500 volts when all the preamp tubes get pulled. Which is probably okay, I'm sure the caps have a safety margin. But to be "perfectly" safe we want to use the calculated values, and we can also split up the bleeder like Marshall does if we need to reduce the wattage on the resistors.
We can verify that. With no downstream draw, that means we'll be dropping 560 volts through 20.2k, which means we'll have 499 volts at the top of the capacitor chain. Check. Current draw here will be 27.7 mA.
At full downstream draw of 20 mA that gives us 47.7 mA total, and therefore the voltage drop through a 2.2k resistor is 104.94 volts, leaving 455 volts at the top of the chain. Check.
So then, wattage through 18k at 500 volts is 13.85 watts, so we can use a 10k and an 8.2k in series at 10 watts each. Problem solved.
Ain't that interesting, these values are almost identical to what Marshall uses in their 200-watter. They use a 1.8k dropper and a somewhat complicated arrangement of 3 sets of bleeders that works out to about 38k. So they're "slightly above" 500 volts when all the preamp tubes get pulled. Which is probably okay, I'm sure the caps have a safety margin. But to be "perfectly" safe we want to use the calculated values, and we can also split up the bleeder like Marshall does if we need to reduce the wattage on the resistors.
- Thread starter
- #132
This amp needs the following 6 voltages in the cap chain:
450 for the driver
420 for the reverb
390 for the PI
375 for the cathode follower
360 for the gain stages
350 for the preamp stages
So we can use 6 x 100k for bleeders on each stage, which will give us 16.6k total when they're in parallel, and it'll be "slightly more" because of the added resistance of the downstream droppers. Perfect.
Each 100k resistor will require at most 3 watts, so we can use 100k 5w which is readily available off the shelf. See how easy that was?
Now we have a perfectly safe amp and we can work on it with or without the tubes installed.
At this point we can begin trimming the dropper resistors, and some small tweaks to the bleeders will likely be beneficial. The current draw decreases dramatically after the PI, so the dropper values will go up after the third one, which means we can get away with smaller (less wattage) bleeders in the preamp section. 5 watt resistors are pretty small, and besides they look pretty and they fit perfectly between the sections of a cap can.
This calculation is essential before turning the amp on. You don't want 500 volt electrolytics blowing up in your face. They stink when they do, the smell won't leave you for days. You'll be an incel for a week lol.
450 for the driver
420 for the reverb
390 for the PI
375 for the cathode follower
360 for the gain stages
350 for the preamp stages
So we can use 6 x 100k for bleeders on each stage, which will give us 16.6k total when they're in parallel, and it'll be "slightly more" because of the added resistance of the downstream droppers. Perfect.
Each 100k resistor will require at most 3 watts, so we can use 100k 5w which is readily available off the shelf. See how easy that was?
Now we have a perfectly safe amp and we can work on it with or without the tubes installed.
At this point we can begin trimming the dropper resistors, and some small tweaks to the bleeders will likely be beneficial. The current draw decreases dramatically after the PI, so the dropper values will go up after the third one, which means we can get away with smaller (less wattage) bleeders in the preamp section. 5 watt resistors are pretty small, and besides they look pretty and they fit perfectly between the sections of a cap can.
This calculation is essential before turning the amp on. You don't want 500 volt electrolytics blowing up in your face. They stink when they do, the smell won't leave you for days. You'll be an incel for a week lol.
toobfreak
Tungsten/Glass Member
- Apr 29, 2017
- 98,938
- 104,937
- 3,615
Probably not an easy option for you, Scruff, but the way I'd tackle that would be to set that plate up on a milling table and run a fly cutter over it. Take that crap right off and give you a perfect surface.
- Thread starter
- #134
Initial estimates then put the droppers as follows:
560 to 450: 1.8k (20 mA)
450 to 420: 1.5k (18 mA)
420 to 390: 2.4k (12 mA)
390 to 375: 1.8k (8 mA)
375 to 360: 3.0k (5 mA)
360 to 350: 5.0k (2 mA)
These are all standard values except for the last two, and we can use 3.3k and 4.7k instead, which will give us about one volt less on the gain stages, which is perfectly acceptable.
Adding up all the values we get 14k, which is about in the same ballpark as Marshall's 10k (give or take). So we'll start there, and adjust as needed.
"Amp parts":
That's a Grayhill rotary switch in the red tube, it's for the cathode caps.
I've tried every coupling cap under the sun, including the $15 red Astrons, to my ear the el cheapo polypropylene caps actually sound the best. Orange Drops are harsh, and the expensive vintage reissues don't add anything of value (and besides they're huge which makes them undesirable in tight spaces).
For filter caps I use the JJ's, they're more reliable than the F&T's. Sprague Atoms for the bias circuit, their 150 volt versions are nice and small. For the presence caps I'm stuck with Mallory 150's, they're the only small .68 out there, they sound fine in that position. For resistors, always metal oxide, carbon comps are unreliable and they drift a lot. I like the Bourns pots but they don't come in all values, so CTS for the others. Anything switched gets a Grayhill whenever possible. What else... 6A10 diodes for the power supply, Hammond transformers (you can get them through Mouser or Hawk or AES), JJ tubes (because NOS RCA's are prohibitively expensive)...
560 to 450: 1.8k (20 mA)
450 to 420: 1.5k (18 mA)
420 to 390: 2.4k (12 mA)
390 to 375: 1.8k (8 mA)
375 to 360: 3.0k (5 mA)
360 to 350: 5.0k (2 mA)
These are all standard values except for the last two, and we can use 3.3k and 4.7k instead, which will give us about one volt less on the gain stages, which is perfectly acceptable.
Adding up all the values we get 14k, which is about in the same ballpark as Marshall's 10k (give or take). So we'll start there, and adjust as needed.
"Amp parts":
That's a Grayhill rotary switch in the red tube, it's for the cathode caps.
I've tried every coupling cap under the sun, including the $15 red Astrons, to my ear the el cheapo polypropylene caps actually sound the best. Orange Drops are harsh, and the expensive vintage reissues don't add anything of value (and besides they're huge which makes them undesirable in tight spaces).
For filter caps I use the JJ's, they're more reliable than the F&T's. Sprague Atoms for the bias circuit, their 150 volt versions are nice and small. For the presence caps I'm stuck with Mallory 150's, they're the only small .68 out there, they sound fine in that position. For resistors, always metal oxide, carbon comps are unreliable and they drift a lot. I like the Bourns pots but they don't come in all values, so CTS for the others. Anything switched gets a Grayhill whenever possible. What else... 6A10 diodes for the power supply, Hammond transformers (you can get them through Mouser or Hawk or AES), JJ tubes (because NOS RCA's are prohibitively expensive)...
~S~- Thread starter
- #136
Beware of shysters selling wire on eBay.
You can get excellent wire for cheap from Mouser or any one of half a dozen other places.
There's a bunch of people on eBay selling 50 year old surplus wire as if it were new.
I'm using top coated stranded wire, it shapes even more easily than solid and it's available for 20 cents a foot or less. All different gauges and colors, whatever you want.
You can get excellent wire for cheap from Mouser or any one of half a dozen other places.
There's a bunch of people on eBay selling 50 year old surplus wire as if it were new.
I'm using top coated stranded wire, it shapes even more easily than solid and it's available for 20 cents a foot or less. All different gauges and colors, whatever you want.
- Thread starter
- #137
All right, I have to go back for another round of parts (forgot a few things last time lol).
So, let's talk about vacuum tubes for a minute.
To make this amp work, it needs a matched pair of KT-88's. And the reality is, most matched pairs, aren't. To get a properly matched pair, you have to pay someone to take the tube out of the box and run it through a mutual conductance tube tester. Apex, for instance, doesn't do this. An "Apex matched pair" just means the tubes have been tested.
What's worked for me in the past, is to buy a quad, and "usually" two of the tubes will be pretty close. Matching means different things to different people. My expectation is, that when I put two tubes into a push pull circuit with a single bias feed, they will read within 10% of each other on the bias meter. This is a different concept from "gain matching", and in a way it's a more stringent requirement.
Vacuum tubes are generally all over the map, no two tubes are alike. The manufacturing process is difficult and chancey. With 12ax7's it doesn't much matter, they're cheap so you can keep trying them till you find one that works. But KT-88's cost near 100 bucks a piece, and most people can't afford to tear up hundred dollar bills. With KT-88's especially, the bias is super important. You have to be between 40 and 60 mils, outside that range the tubes become very unhappy. In a push pull amp circuit, if you have one tube sitting at 47 and the other at 53, that's not too bad. (They'll drift over time anyway, if you're using the amp every day you should check every 3 months).
When bringing up the amp, the first thing I do is finish the heater wiring and install the tubes and make sure they light up. The next thing is apply power to the output tubes and make sure they bias correctly. (This is the first part of making the amp DC-happy, and it's very easy, you don't need anything but the power supply and the output transformer).
To bias the amp, start with the bias control all the way to the right (maximum bias voltage), and turn the amp on and take it out of standby. My maximum bias voltage is about -100 volts, so when I turn the amp on I see a "small" current, maybe 20 mils or less. Then back off on the bias control till the hottest tube reads 50 mA. If the other one reads less than 40 you have mismatched tubes and you need to try a different pair. If they're pretty close to each other, back off on the bias control some more, till the average reading of the two tubes is 50. This is the correct bias point.
You must then LOCK DOWN the bias control with some glypt or something. Never put the bias control on the back panel where it can move or get bumped, always put it inside the chassis and use a pot that can only be adjusted with a screwdriver. Use only a LITTLE glypt (a single drop is usually enough), because you'll have to adjust again whenever you replace the tubes.
Once you're done, LABEL which tube is which, because if you're going to work on the amp some more you'll have to take them out again. Always use 1% bias resistors, and never use carbon comps, always use metal oxide. Bias current through these resistors is tiny, you don't need big ones.
When I do this for the first time on a new amp, after the initial bias adjustment, I let the tubes sit for about 15 minutes with the power on. Sometimes they drift a little, if they do you can readjust before applying glypt. After 15 minutes they should be totally stable, if they're not start over with a different pair of tubes.
Emphatically, the tubes need to be within 10% of each other. For 50 mils that means your range is 47-53. If they're outside that range you need a better match.
If you're matching tubes yourself using a Hickok 600a or some other mutual conductance tester, you want your bias meter in circuit while you're performing the test. Mutual conductance measures gain, using a fixed bias. With the bias meter in and IF you're lucky, you can match for both gain and bias at the same time.
My experience has been that whenever you buy an Apex matched quad, you'll have to throw one of the tubes away. (Or sell it on eBay lol). Three of them will be 45-55 and the fourth one will be 11 or something. So use the two closest ones and then you'll have a spare for emergencies. In an emergency (like if you blow a tube during a gig) you can run with the spare even if it's 20% unmatched, but the minute the gig is over you should go buy some more tubes. 20% unmatched at 50 mils gives you a range of 40 to 60, and you should absolutely NOT use the amp outside of that range.
This is one very good reason to always put an effects loop into your amp. In an emergency when you blow a power tube, you can give the effects send to the sound guy and he can run you through the board. (Effects send or "preamp out", either way is fine - in the case of preamp out you want to be able to disconnect your power amp when the plug is in, which is something different from simply tapping the preamp into the board).
Do not, EVER, run the amp with a brand new untested tube or tubes. Always adjust the bias first and make sure it's in range. Usually this is the job of the guitar tech but for most people you're your own tech, so be conservative and don't take risks. Even a hundred dollar tube is cheaper than a two thousand dollar amp, and playing one gig through the board isn't going to kill you.
So, let's talk about vacuum tubes for a minute.
To make this amp work, it needs a matched pair of KT-88's. And the reality is, most matched pairs, aren't. To get a properly matched pair, you have to pay someone to take the tube out of the box and run it through a mutual conductance tube tester. Apex, for instance, doesn't do this. An "Apex matched pair" just means the tubes have been tested.
What's worked for me in the past, is to buy a quad, and "usually" two of the tubes will be pretty close. Matching means different things to different people. My expectation is, that when I put two tubes into a push pull circuit with a single bias feed, they will read within 10% of each other on the bias meter. This is a different concept from "gain matching", and in a way it's a more stringent requirement.
Vacuum tubes are generally all over the map, no two tubes are alike. The manufacturing process is difficult and chancey. With 12ax7's it doesn't much matter, they're cheap so you can keep trying them till you find one that works. But KT-88's cost near 100 bucks a piece, and most people can't afford to tear up hundred dollar bills. With KT-88's especially, the bias is super important. You have to be between 40 and 60 mils, outside that range the tubes become very unhappy. In a push pull amp circuit, if you have one tube sitting at 47 and the other at 53, that's not too bad. (They'll drift over time anyway, if you're using the amp every day you should check every 3 months).
When bringing up the amp, the first thing I do is finish the heater wiring and install the tubes and make sure they light up. The next thing is apply power to the output tubes and make sure they bias correctly. (This is the first part of making the amp DC-happy, and it's very easy, you don't need anything but the power supply and the output transformer).
To bias the amp, start with the bias control all the way to the right (maximum bias voltage), and turn the amp on and take it out of standby. My maximum bias voltage is about -100 volts, so when I turn the amp on I see a "small" current, maybe 20 mils or less. Then back off on the bias control till the hottest tube reads 50 mA. If the other one reads less than 40 you have mismatched tubes and you need to try a different pair. If they're pretty close to each other, back off on the bias control some more, till the average reading of the two tubes is 50. This is the correct bias point.
You must then LOCK DOWN the bias control with some glypt or something. Never put the bias control on the back panel where it can move or get bumped, always put it inside the chassis and use a pot that can only be adjusted with a screwdriver. Use only a LITTLE glypt (a single drop is usually enough), because you'll have to adjust again whenever you replace the tubes.
Once you're done, LABEL which tube is which, because if you're going to work on the amp some more you'll have to take them out again. Always use 1% bias resistors, and never use carbon comps, always use metal oxide. Bias current through these resistors is tiny, you don't need big ones.
When I do this for the first time on a new amp, after the initial bias adjustment, I let the tubes sit for about 15 minutes with the power on. Sometimes they drift a little, if they do you can readjust before applying glypt. After 15 minutes they should be totally stable, if they're not start over with a different pair of tubes.
Emphatically, the tubes need to be within 10% of each other. For 50 mils that means your range is 47-53. If they're outside that range you need a better match.
If you're matching tubes yourself using a Hickok 600a or some other mutual conductance tester, you want your bias meter in circuit while you're performing the test. Mutual conductance measures gain, using a fixed bias. With the bias meter in and IF you're lucky, you can match for both gain and bias at the same time.
My experience has been that whenever you buy an Apex matched quad, you'll have to throw one of the tubes away. (Or sell it on eBay lol). Three of them will be 45-55 and the fourth one will be 11 or something. So use the two closest ones and then you'll have a spare for emergencies. In an emergency (like if you blow a tube during a gig) you can run with the spare even if it's 20% unmatched, but the minute the gig is over you should go buy some more tubes. 20% unmatched at 50 mils gives you a range of 40 to 60, and you should absolutely NOT use the amp outside of that range.
This is one very good reason to always put an effects loop into your amp. In an emergency when you blow a power tube, you can give the effects send to the sound guy and he can run you through the board. (Effects send or "preamp out", either way is fine - in the case of preamp out you want to be able to disconnect your power amp when the plug is in, which is something different from simply tapping the preamp into the board).
Do not, EVER, run the amp with a brand new untested tube or tubes. Always adjust the bias first and make sure it's in range. Usually this is the job of the guitar tech but for most people you're your own tech, so be conservative and don't take risks. Even a hundred dollar tube is cheaper than a two thousand dollar amp, and playing one gig through the board isn't going to kill you.
man, i gotta lot to learn here Scruff.......i found tubes for my '66 SUNN bass head......ignorantly installed them and cranked it up......
~S~
- Thread starter
- #139
man, i gotta lot to learn here Scruff.......i found tubes for my '66 SUNN bass head......ignorantly installed them and cranked it up......
View attachment 1152131
~S~
The 200s runs at lower voltage, it's rated at 60-65 watts and has a B+ of 480 volts. The nominal bias is -55 volts.
The bias control is right between the KT-88 output tubes, you can see it in this pic, to the left of the blue filter caps. (Gotta love those Black Cats!)
Here is the schematic:
The output section is similar to mine, you should strive for a matched pair of power tubes if possible. IMHO.
At my slightly higher plate voltage of 560 volts, 60 mA of cathode current takes the tube to 70% of its dissipation limit, which is probably the highest you want to run it. 75 mA per tube exceeds the dissipation limit, the plates will start to glow red after that.
With 480 volts you have more headroom, you can use 66 mA of bias current and still stay well within the guidelines. This is assuming you have KT-88's, sometimes a 200s has 6550's instead, which is a different animal. (Don't believe the hype about them being the same, they're NOT the same, a 6550 has a considerably lower dissipation limit).
The 200s is a great amp. If you haven't done so already, I'd consider replacing the ground switch (and cap) with a properly grounded 3-wire AC cord. That's an electrical safety thing, it won't affect the sound. Ground caps are dangerous.
I hope it goes without saying, never randomly tweak the bias pot, there's no cathode resistor in these amps so you'll need a bias gadget.
If it works and it sounds good and the plates aren't glowing cherry red you're probably okay, but if it were me I'd check it anyway to make sure.
What brand of output tubes are you using?
Similar threads
New Topics
-
Roberts on personal attacks against judges: 'It's got to stop'- Started by C_Clayton_Jones
- Replies: 1
-
I give Al Jazeera credit for publishing this opinion piece: 'The US-Israeli Strategy Against Iran is Working. Here is Why'- Started by DigitalDrifter
- Replies: 7
-
-
