Need help troubleshooting Curtis Mathes conversion re-do.

Hello! I've been lurking on this fine forum for some time and have finally come upon a reason to join. I've recently begun re-doing a previously converted Curtis Mathes console power amp which I picked up on ebay and has been amplifying my son's ipod for the past few years.

I have been unable to identify the model of this amp or locate a schematic, but the transformer date is 1960 and it seemed like a pretty straight forward push-pull power amp. It is tube rectified with a 5U4 and there is a single RCA 7025 driver/inverter, (which is a 12AX7 variant and turns out to be a very valuable tube indeed. Tube Depot sells them for $180 ea.!) and four SBQ5/el84 output tubes.

My goals were:
  • clean up the wiring
  • relocate the inputs and outputs that were for some reason located on the power supply side
  • swap the positions of the rectifier tube and can filter capacitors
  • replace the radial electrolytics soldered to the old can with a new can and preamp filter
  • add a volume control and more substantial power/standby combo switch
  • add an external fuse holder.
  • add heater supplied pilot light

There was nothing wrong with how the amp operated, that is, until after I started monkeying with it. The issues are to follow, but first some pics!

Here is what it looked like before...

QS6VTQa.jpg


As you can see, lots of unnecessary open-ended wiring originally connected to the preamp/tuner socket. Also both the input jacks and speaker binding posts are located next to the power supply. Another odd thing (odd to me anyway) is that one side of the 5V winding for the rectifier heaters was connected to the first filter cap BEFORE the rectifier tube. I'm sure there is a sound reason for this, which probably has to do with quieting hum, but it's not something I am familiar with.

Here is what the amp looks like now...

hbXvyiP.jpg


zPj3QqX.jpg


W0RznCJ.jpg



For a short time after the aforementioned changes the amp seemed to operate normally and idled very quietly with almost no discernible hum. I had it connected to an 8 ohm speaker load as it has been for years (although I think the output wants a 4 ohm load) and the new volume control worked perfectly. Gradually (after 30 minutes or so) the output reduced to barely a whisper. No distortion, just very little output. Also both the chassis and the power transformer felt unusually hot.

I opened the amp back up to check voltages and nothing seemed out of the norm compared to it's previous incarnation's voltages. B+, plates, grids and cathodes were unchanged on both preamp and output stages.

Here's what seems very strange:
I accidentally checked one of the heater pins on an output tube with my voltmeter set on DCV and it registered voltage. I thought I was mistaken and set the meter to VAC and it did read voltage but unusual numbers. One winding was putting out 7.5 VAC and the other 2.8 VAC. I then switched the meter back to DCV and it did actually read 4 DCV on one side and 3.6 DCV on the other side. This is not something I have seen before. The un-rectified, un-filtered heater windings are coming straight out of the PT and somehow putting out voltage that is both AC and DC. Can they be registering both AC and DC volts simultaneously? Also, I wonder if this higher voltage may prematurely burn out the tube filaments. Now, checking the voltage there it reads 4 VAC and 2.8 VAC, so fluctuating quite a bit.

Aside from adding a volume control and deleting a bunch of open-ended hookup wire, the only alterations I made to the previous circuit were
  • replaced the pair of main filter caps from 68uf to 50uf (the original can had a 60 and 30 uf cap in it). I'll admit there were no total current draw calculations involved there, I just used what I had that was in the neighborhood.
  • changed the preamp filter cap from 100uf to 33uf
  • replaced one cathode bypass cap of same value.
  • Moved the 5V winding leg to the rectifier before the first main filter cap.

Here are the readings currently:

PT
HT Standby 325 VAC
On 311 VAC
6.3V heater windings 4.0, 2.8 VAC
5V windings
without 5u4 10.6/6.9 VAC
with 5U4 334 B+ DCV

V1
pin 1 (plate 1) 304
pin 2 (grid 1) .9 DCV (at full volume)
pin 3 (cathode) 5.0
pin 6 (plate 2) 304
pin 7 (grid 2) .9 DCV
pin 8 (cathode) 4.90

V2
plate 318
grid 314
cathode 11.0

V3
plate 321
grid 314
cathode 11.0

V4
plate 321
grid 314
cathode 10.1

V5
plate 321
grid 314
cathode 10.1

Any comments, ideas or reprimands are much appreciated.

~ C
 
Last edited:
Here is a good thread with an updated schematic on your amp.

http://audiokarma.org/forums/index....to-do-with-this-curtis-mathes-console.603803/

If you are measuring the filaments to ground, the hum pot, if used may not be centered.
You should replace the original coupling caps.
If the lug in the center of the chassis is grounded the black cap looks like it may be in backwards.
If it is insulated and fed from the can cap it is correct.
The symptoms sound like bad coupling caps, or an electrolytic cap is backwards
 
Last edited:
The thread Scott referenced above of my Curtis Mathes amp appears to be a slightly different model than yours. For example the 12AX7 is located behind the power tubes on yours, whereas on mine, it is located in front of the power tubes. My original power amp used the 12AX7 as a split load (cathodyne) inverter/driver and in that configuration, provided no stage gain at all. Because of that, there wasn't sufficient voltage gain to drive the amp directly from a line source. To address that, I needed to add an extra gain stage. This gain stage was originally located on the preamp chassis. So I moved it to the amp chassis and then added a volume control in front of that gain stage. So that is what the schematic shows on my thread.

I am unclear how your amp is obtaining enough gain to be used as a standalone "integrated" amp since there is only one 12AX7 tube present. The function of that tube must be for phase inverter duties for both channels, therefore it must be configured as a split load (cathodyne) type phase inverter. But perhaps your son's iPod was putting out enough voltage to overcome the lack of gain in the power amp.

Some ideas to mull over:
1. When you measure AC voltage you need to measure across the load, not between a certain point and ground. In other words, to see how much voltage is really present on the filament winding, you need to place the volt meter probes across pins 4 and 5 of one of the EL84 tubes.
2. There was a wiring error on my amp with the 5V filament winding. One side of the 5V filament winding was connected directly to the positive pin of the first filter cap. This is incorrect. The correct method is to connect one side of the 5V filament winding directly to pin 2 and the other side directly to pin 8. Thus the heater current makes a complete circuit directly around the filament winding. A tap wire is then connected from pin 8 of the rectifier tube to the positive terminal of the first filter cap, which then carries only the current needed to run the amp. Not sure if your amp has this same wiring error or not.
3. The voltages you measured on pins 1 & 3 and pins 6 & 8 of the 12AX7 appear to be biasing each of those triodes into cutoff condition. This would certainly explain why there is no sound coming out of the amp. Evidence of this is not only that the difference in voltage between pins 2 and 3 is something like -4V (should be at most maybe -2.5V), but that almost all of the power supply rail voltage is seen at pins 1 and 6, which indicates no current is flowing across those plate resistors and thus the tube itself is taking all of the voltage drop (about 300V, between plate and cathode). This has got to be the crux of your problem. See if you can find out what is causing the two 12AX7 triodes to be biased into cutoff. Fix that, and your amp will likely start working again.
4. I can't see it in the picture you provided but there needs to be a DC blocking capacitor on the input of a self biased stage (C1 in the diagram below), if that stage is taking a feed directly from an external line source. Maybe it is there but obscured from the angle of the camera in the photo. If that capacitor is not there, it could skew the biasing of that stage when an external source is connected.
5. Also in a typical cathodyne stage that I believe is being used in this amp, RL and RK should be the same value, and for a 12AX7 cathodyne stage, Rb should be somewhere between maybe 800R and 3K.

As a reminder, when measuring voltages on a self biased stage with a large Rg (like 470K to 1M), a typical digital volt meter will not measure grid voltage accurately. To measure grid voltage in this setup, you need to measure between the cathode and ground, and then again at the junction of Rb, Rg, and RK and ground, and then subtract those two voltages (or just measure the voltage drop directly across Rb) to arrive at a correct value for the bias voltage. For a properly biased 12AX7 cathodyne stage, the grid voltage should be somewhere between maybe -0.7 and -2.5V below the cathode voltage.

upload_2017-1-16_11-48-48.png

Kevin
 
Last edited:
Just FYI, a 7025 is just a slightly fancier 12AX7. You can replace it with a 12AX7 and it'll be fine. You can get 7025 tubes on eBay easily for $10-20, though obviously some higher-end ones go higher.
 
Its actually a low noise premium 7025. Absolutely functionally the same as a 12ax7 though, same socket and all.

Got a schematic for this? It would help tremendously.

The AC and DC on the heater string isn't really unusual. A lot of amplifiers connected the center tap or the hum pot to the cathode of the output tube. This puts some DC on the heater for quieting reasons, and obviously still has the AC to actually light the heaters. You'd need to measure across the heater winding on AC to see what is lighting them, reading either to ground on DC would give you cathode voltage or whatever DC source voltage they tapped. Once in a while its a resistor off the HV circuit.

Functionally it doesn't much matter whether the 5v heater leg lands on the rectifier or the filter cap first. The two parts will be connected together anyway. The bigger question is where the HV center tap lands. Ideally it ought to land on the negative terminal of the filter cap. This makes things as quiet as its inclined to be. Some amps didn't do this and managed to be quiet anyway, but if I'm sticking my grubby little hands inside one and can change this simply by extending the wire, I do so.

Definitely agree that it sounds like dead coupling caps. Those black beauties are known to get leaky after a while.


With no negative feedback loop, it might actually have enough gain to work. Fair guess that the original setup had a tie from the speaker jack up through the preamp where it tied into the voltage amplifier tube to close the loop. Personally I'd dig up the preamp schematic to confirm, and add another tube to the chassis to duplicate the original configuration with a voltage amp and a phase inverter, along with closing the NFB loop.
 
My amp had a small 22 gauge wire connecting pin 8 of the rectifier to the first filter cap, where the other side of the 5V winding (with its 18 gauge wire) was connected. So that small gauge wire was carrying the full rectifier heater voltage. That didn't seem "proper" to me.

Yes, on my amp (no schematic but I reverse traced it), there was in fact a feedback wire from the secondary of the OPT back through the umbilical to the preamp chassis that was connected to the cathode circuit of the other 12AX7 used as the voltage gain stage.
 
Thanks so much for all of the thoughtful comments and suggestion!. I'm spending some time with the amp this evening. Yes, an original schematic would be soooo helpful. Just knowing the model of the amp would be a start.

Measuring the voltage across the heater windings, as suggested, rather than to chassis ground, yielded 7+ DCV, so I'm relieved there is normal voltage there. Most of the amps I've worked with have been guitar amplifiers which all have had three-prong cords and grounded AC voltage. This amp has a two-prong cord, so perhaps this is the difference when measuring the heaters. By the way, would it be a no-no to add a three-prong power cord to this amp???

While writing this post I decided to hook the amp back up to speakers and my turntable and it is once again amplifying the signal, albeit not very loud. Not sure why. It's a nice clean signal and tone and very quiet noise/hum-wise but it sounds like a 5 watt amp and not even as loud as a little single ended Magnavox stereo amp the turntable is normally plugged into. I'll leave it playing for a while to see if the volume drops off again.

SWTlCkr.jpg


I'm sure Kevin is correct that a gain stage needs to be added in order to use it as an integrated amplifier and that my son's ipod was acting as a small preamp to boost the signal. I may copy Kevin's design from the schematic and add another preamp tube. I'd like to be able to plug a turntable into it and have some decent volume.

As far as the input DC blocking caps go, well there weren't ones there to begin with before I added the volume pot so I figured they weren't intended to be there. It doesn't appear that the original preamp or output circuits have been altered at all so I'm assuming they weren't part of the original design, although I know it's a bit unconventional. Sound and hum-wise it doesn't appear to need them.

As far as the coupling and signal caps go, they all "seem" to be functioning well. I've had a lot of vintage guitar amps from the 50's and 60's and unless a cap is clearly showing signs of failing I usually leave it alone. Electrolytics are another story as I know that if they fail they can do irreversible harm to trannies and power tubes. I did replace all of the electrolytics with the exception of one cathode bypass cap.

The amp is still amplifying nearly an hour later. The transformer is quite warm to the touch but I'm assuming this is normal based upon the cooling fins that were originally on these amps. The person who did the original conversion on this amp must have removed them.
 
I forgot to mention that I paralleled two 8ohm speakers on each side of the output to meet the 4ohm load and this seemed to increase the output by a margin.
 
Coupling caps that leak, cause the output tubes to lose bias, and cause the tubes to draw more current, which could ruin the tubes,output transformers(if a tube fails)or the power transformer from overheating.It is not always obvious when they are leaking until the output tubes start redplating.The coupling caps are very cheap.
 
The heater would need to be read in AC volts, not DC.

You'd need a phono preamp to go with this. Phono cartridge output is around 3 mv, most amplifiers have an input sensitivity closer to 1 volt in order to get full output. No EQ curve either. The ceramic cartridge was probably close to 1 volt output but those tend towards horrible.
 
I'm trying to remember how my preamp was connected to the power amp originally. The coupling caps may have been provided in the preamp, or it may have been direct (DC) coupled. The few remaining pictures I have of it aren't that helpful in determining that. In any case, it sure makes sense to me to add coupling caps if you want to continue to use the amp as is, or rebuild it to add another gain section like I did on mine.
 
Coupling caps that leak, cause the output tubes to lose bias, and cause the tubes to draw more current, which could ruin the tubes,output transformers(if a tube fails)or the power transformer from overheating.It is not always obvious when they are leaking until the output tubes start redplating.The coupling caps are very cheap.

Ah. Are you referring to the PI coupling capacitors at the output tube grids?

The heater would need to be read in AC volts, not DC.

You'd need a phono preamp to go with this. Phono cartridge output is around 3 mv, most amplifiers have an input sensitivity closer to 1 volt in order to get full output. No EQ curve either. The ceramic cartridge was probably close to 1 volt output but those tend towards horrible.

Sorry I meant AC volts!

Can you recommend a good phono preamp design that I could adapt to this amp? Would that be like shooting in the dark without a schematic?
 
I'm trying to remember how my preamp was connected to the power amp originally. The coupling caps may have been provided in the preamp, or it may have been direct (DC) coupled. The few remaining pictures I have of it aren't that helpful in determining that. In any case, it sure makes sense to me to add coupling caps if you want to continue to use the amp as is, or rebuild it to add another gain section like I did on mine.
I guess it would make sense especially if that first remote preamp tube was used as the input tube.
 
I found a great preamp design to add to the front end of this amp. I believe this was from the diy audio projects site:

OmVwsVy.jpg


It turned out to be a really great sounding addition to this little power amp. I added two more 12AX7's and the circuit fit in quite nicely between the newly added volume pot and cathodyne phase inverter. With the exception of adjusting one of the B+ dropping resistors, I used this exact design and plugged it's output into the PI grids via two .047 coupling caps. I also ended up replacing all of the 9 pin sockets because the old ones were, well, old and also most of the hookup wire as well.

I played just about every style of music on my ipod through it at full volume with a set of 8 ohm bookshelf speakers and the amp has waaay more gain and a much fuller, dynamic sound with no discernible distortion, hum, etc..I'm extremely pleased with the result.

I also discovered that my turntable has a built-in preamp, so really no need for a phono preamp.

Here's the final version:

6sQ9sVn.jpg


jkntKqB.jpg


2A5wrPH.jpg


I am now interested in finding an excellent set of 4 ohm high efficiency bookshelf speakers with a high db rating. Any suggestions out there???
 
Now you're cooking with gas.

One additional mod you may want to consider if you feel like tearing into it again is to add some global feedback that encompasses the output transformer, i.e., taken from the secondary of the output transformer back to the second triode in the "preamp" section. (I put preamp in quotes because according to the original Curtis Mathes design, the second stage of the preamp is really the voltage gain stage of the power amp.) You typically would not put a volume pot in between the voltage gain stage and phase inverter/driver stage of a power amp because that would put the volume pot inside the feedback loop. Having a volume pot inside a feedback loop will essentially change the amount of feedback applied based on volume pot setting.

In any case, adding some global feedback will provide numerous benefits such as extend frequency response, allow the amp to react better to high frequency transients, lower distortion and noise (not that noise is a problem in your case), as well as lower power amp output impedance (meaning increase damping factor). Overall it will give you a better sounding amp if you apply the feedback judiciously. (That's my opinion, not everyone may agree with that statement.) If you look at my schematic on my Curtis Mathes build thread, you will see that is exactly what I did. By way of information this is also the basic design of the original Curtis Mathes amp (even though the Curtis Mathes engineers located the voltage gain stage of the power amp on the preamp chassis).
 
Now you're cooking with gas.

One additional mod you may want to consider if you feel like tearing into it again is to add some global feedback that encompasses the output transformer, i.e., taken from the secondary of the output transformer back to the second triode in the "preamp" section. (I put preamp in quotes because according to the original Curtis Mathes design, the second stage of the preamp is really the voltage gain stage of the power amp.) You typically would not put a volume pot in between the voltage gain stage and phase inverter/driver stage of a power amp because that would put the volume pot inside the feedback loop. Having a volume pot inside a feedback loop will essentially change the amount of feedback applied based on volume pot setting.

In any case, adding some global feedback will provide numerous benefits such as extend frequency response, allow the amp to react better to high frequency transients, lower distortion and noise (not that noise is a problem in your case), as well as lower power amp output impedance (meaning increase damping factor). Overall it will give you a better sounding amp if you apply the feedback judiciously. (That's my opinion, not everyone may agree with that statement.) If you look at my schematic on my Curtis Mathes build thread, you will see that is exactly what I did. By way of information this is also the basic design of the original Curtis Mathes amp (even though the Curtis Mathes engineers located the voltage gain stage of the power amp on the preamp chassis).

Kevin
Thanks for the suggestions. I have the line inputs connected to the left lugs of the volume pot and then the center lugs connected to the input of the preamp tubes. Then the output of each of those goes to the grids of the PI. Does that sound correct? It seems to work well anyway.

I will try the global feedback arrangement. Always willing to tinker.

How did you determine that the output impedance (if that is the correct terminology) is 4 ohms for this amp? Just curious...
 
Yes that volume pot wiring sounds correct.

Concerning the output transformer secondary impedance:

1. First I made an assumption that the Curtis Mathes engineers would have specified a primary impedance for optimal loading, which would be around 8K. I then measured the turns ratio of the transformer (by feeding a sine wave into the primary at a known amplitude and measuring the amplitude on the secondary). The voltage ratio measured (assuming an 8K-ish primary) indicated that the secondary was right around 4 ohms.
2. After the amp was built, I hooked up both an 8 ohm resistive load and then a 4 ohm resistive load and drove the amp with a 1 KHz sine wave to just before clipping in both cases, as observed on an oscilliscope. With the 8 ohm load, the amp delivered about 7 watts output (as I recall), but under a 4 ohm load, it delivered about 11.5 watts output before clipping.

Knowing from previous experience that an EL84 push pull output stage configured similarly as the Curtis Mathes amp should be capable of delivering about 12 watts output if primary is optimally loaded, it led to the conclusion that the secondary was indeed 4 ohms.

3. As a final sanity check, under a 4 ohm dummy load I measured the low frequency extension at full power (about 11.5 watts), and found that to be right at 50 Hz before the signal started to fall apart. Given the relative small size of the output transformers, and seeing that the amp was capable of reproducing 11.5 watts down to 50 Hz is another bit of evidence that there is enough magnetic flux provided in the output transformer iron to support the conclusion of an 8K to 4 ohm transformer.

But...because the turns ratio is relative with respect to primary vs secondary impedances, I can't be 100% certain. The three pieces of evidence however do stack up in favor of that conclusion.

By the way, for my amp, I found some Advent II bookshelf speakers off the auction site that are 4 ohm. They needed to be refoamed but they sounded pretty good. Not as good as my modern bookshelf speakers, but not bad.
 
Working out the winding ratios and figuring out what seems appropriate is usually how its done. Back it up with some performance testing at various loads to confirm. I have one that works out to want a 2.5 ohm load for best performance. I haven't done anything with that amp other than steal it's tubes for other purposes.
 
Yes that volume pot wiring sounds correct.

Concerning the output transformer secondary impedance:

1. First I made an assumption that the Curtis Mathes engineers would have specified a primary impedance for optimal loading, which would be around 8K. I then measured the turns ratio of the transformer (by feeding a sine wave into the primary at a known amplitude and measuring the amplitude on the secondary). The voltage ratio measured (assuming an 8K-ish primary) indicated that the secondary was right around 4 ohms.
2. After the amp was built, I hooked up both an 8 ohm resistive load and then a 4 ohm resistive load and drove the amp with a 1 KHz sine wave to just before clipping in both cases, as observed on an oscilliscope. With the 8 ohm load, the amp delivered about 7 watts output (as I recall), but under a 4 ohm load, it delivered about 11.5 watts output before clipping.

Knowing from previous experience that an EL84 push pull output stage configured similarly as the Curtis Mathes amp should be capable of delivering about 12 watts output if primary is optimally loaded, it led to the conclusion that the secondary was indeed 4 ohms.

3. As a final sanity check, under a 4 ohm dummy load I measured the low frequency extension at full power (about 11.5 watts), and found that to be right at 50 Hz before the signal started to fall apart. Given the relative small size of the output transformers, and seeing that the amp was capable of reproducing 11.5 watts down to 50 Hz is another bit of evidence that there is enough magnetic flux provided in the output transformer iron to support the conclusion of an 8K to 4 ohm transformer.

But...because the turns ratio is relative with respect to primary vs secondary impedances, I can't be 100% certain. The three pieces of evidence however do stack up in favor of that conclusion.

By the way, for my amp, I found some Advent II bookshelf speakers off the auction site that are 4 ohm. They needed to be refoamed but they sounded pretty good. Not as good as my modern bookshelf speakers, but not bad.

Thanks, Kevin. Clearly I asked the right person. What a great resource you are for this obscure amp project.
 
Back
Top Bottom