Need MC225 help

[Joe Reeder]

I am trying to refurbish my mc225. It's been around the block a few times, being hauled around Texas where I worked until retirement, and in the studio and on the road. Most of the voltages are within 10%. In fact, both plates of the 7025 which is used in the place of the circuit's 12AX7, are at 116 volts, which is only .8% high. The -21 volt grid biases rage from -20.8 to -20.9 volts, which puts it within an average 1% tolerance. Other plates are higher than called for. The V4 and V5 12BH7 plates are supposed to be 285 volts, but are 315 and 326 volts respectively, making their average 12% high. The plates on the 12AU7s max out at 10% high and the plates on the 7591 are within 7% as well as the 4-8 screen voltages.

The resistances on pins 2 of the 12AU7s averaged 32% low. Both plate resistances on the 7025 were 134K, which is 26% low. The maximum plate resistances on the 12AU7s after capacitor settling were within 1% as were the pin 6 voltages. The average resistance of all 4 12BH7 plates was within 5%. The average resistances for the 7591 plates were 5% high and for pins 4 and 8 average was 3% high. The average resistance for pin 5 on the 7591s was .8% high.

For comparison, the resistance (not impedance) of the 70V leads of both output transformers were measured and found to be within 10% of each other. The 4, 8, and 16-ohm lines were also checked and found to be identical.

The problem shows itself with the outputs. A 1KHZ squarewave with a 10-ohm 25-watt resistor on each 8 ohm output produces only 5 volts peak to peak on the scope. It's a good, sharp square wave, but at that level, it should be. The resistors used to get warm in the good old days. Less than one-watt rms might work just fine for some headphones but is pretty lousy for speakers. The other clue is that with no signal pins 5 of the 7591s range from .2 to .67 volts with an average of only .385 volts for all 4 when there should be .8 volts at pin 5 of each.

I have replaced all of the filter capacitors, along with all components of the bias circuits and changed the chokes to 68-ohm resistors, and replaced input and inter-stage coupling capacitors and the two cathode capacitors for the 7025. I have also tested the 7591 pin resistances for shorts or opens and found none.

In short, help! It's obvious I'm missing something. I feel like I did many years ago when I was a technician for a company in Odessa, Texas many years ago. I would be going nuts working on a mini computer and my boss Woody would come back, touch a thing or two, and the thing would start working. He would note that his fingers never left his hands. Is there another Woody in the house?

I want to sell this thing I used to like on eBay to raise money for a school fund for the kids of the kids I have had in ministry for the last 14 years. All I need is $2200.00 which a 79-year-old living on Social Security just can't do.

 

[TriodeLuvr]

Have you checked power supply voltages? Modern line voltage tends to be considerably higher than when thais unit was manufactured. Also, being as you have a scope, input a 1kHz sine wave and turn up the generator until the wave begins to round over or clip. The voltage at that point will tell you how much power is being output. You would do this only with a load attached, of course.

Jack

 

[nerdorama]

This might seem like stupid questions, but do you have the level control turned all the way up? What is your source? Can it put out enough voltage to drive the Mac to full power?

 

[w3jn]

Higher than normal plate voltage indicates the tubes are weak and not drawing the specified current at their operating point. Other factors could influence this such as tube bias, plate or cathode resistor drift, but if your plate resistors are within spec your 12BH7s are supposed to be drawing about 5 ma and they're only drawing about 3 ma each. Your 7591s should idle at 20 (.8V/40 ohms) mA but they're idling at 5-10 mA, which if the bias is correct, they're pretty much goners.

A tube tester might help nail this down, but trying new (or better) tubes would prove things one way or the other.

 

[gadget73]

Tired tubes was my thinking as well. Worth having them tested if you don't have known good subs.

 

[ramjet]

I am trying to refurbish my mc225. It's been around the block a few times, being hauled around Texas where I worked until retirement, and in the studio and on the road. Most of the voltages are within 10%. In fact, both plates of the 7025 which is used in the place of the circuit's 12AX7, are at 116 volts, which is only .8% high. The -21 volt grid biases rage from -20.8 to -20.9 volts, which puts it within an average 1% tolerance. Other plates are higher than called for. The V4 and V5 12BH7 plates are supposed to be 285 volts, but are 315 and 326 volts respectively, making their average 12% high. The plates on the 12AU7s max out at 10% high and the plates on the 7591 are within 7% as well as the 4-8 screen voltages.

The resistances on pins 2 of the 12AU7s averaged 32% low. Both plate resistances on the 7025 were 134K, which is 26% low. The maximum plate resistances on the 12AU7s after capacitor settling were within 1% as were the pin 6 voltages. The average resistance of all 4 12BH7 plates was within 5%. The average resistances for the 7591 plates were 5% high and for pins 4 and 8 average was 3% high. The average resistance for pin 5 on the 7591s was .8% high.

For comparison, the resistance (not impedance) of the 70V leads of both output transformers were measured and found to be within 10% of each other. The 4, 8, and 16-ohm lines were also checked and found to be identical.

The problem shows itself with the outputs. A 1KHZ squarewave with a 10-ohm 25-watt resistor on each 8 ohm output produces only 5 volts peak to peak on the scope. It's a good, sharp square wave, but at that level, it should be. The resistors used to get warm in the good old days. Less than one-watt rms might work just fine for some headphones but is pretty lousy for speakers. The other clue is that with no signal pins 5 of the 7591s range from .2 to .67 volts with an average of only .385 volts for all 4 when there should be .8 volts at pin 5 of each.

I have replaced all of the filter capacitors, along with all components of the bias circuits and changed the chokes to 68-ohm resistors, and replaced input and inter-stage coupling capacitors and the two cathode capacitors for the 7025. I have also tested the 7591 pin resistances for shorts or opens and found none.

In short, help! It's obvious I'm missing something. I feel like I did many years ago when I was a technician for a company in Odessa, Texas many years ago. I would be going nuts working on a mini computer and my boss Woody would come back, touch a thing or two, and the thing would start working. He would note that his fingers never left his hands. Is there another Woody in the house?

I want to sell this thing I used to like on eBay to raise money for a school fund for the kids of the kids I have had in ministry for the last 14 years. All I need is $2200.00 which a 79-year-old living on Social Security just can't do.

I agree with w3jn; everything you described are symptoms of weak tubes. The voltages given in the Mc225 assume factory new tubes as that is what was installed by McIntosh when they wrote the service manual. Finally, you'll get more McIntosh specific assistance in the McIntosh forum.

 

[mcgjohn]

I am trying to refurbish my mc225. It's been around the block a few times, being hauled around Texas where I worked until retirement, and in the studio and on the road. Most of the voltages are within 10%. In fact, both plates of the 7025 which is used in the place of the circuit's 12AX7, are at 116 volts, which is only .8% high. The -21 volt grid biases rage from -20.8 to -20.9 volts, which puts it within an average 1% tolerance. Other plates are higher than called for. The V4 and V5 12BH7 plates are supposed to be 285 volts, but are 315 and 326 volts respectively, making their average 12% high. The plates on the 12AU7s max out at 10% high and the plates on the 7591 are within 7% as well as the 4-8 screen voltages.

The resistances on pins 2 of the 12AU7s averaged 32% low. Both plate resistances on the 7025 were 134K, which is 26% low. The maximum plate resistances on the 12AU7s after capacitor settling were within 1% as were the pin 6 voltages. The average resistance of all 4 12BH7 plates was within 5%. The average resistances for the 7591 plates were 5% high and for pins 4 and 8 average was 3% high. The average resistance for pin 5 on the 7591s was .8% high.

For comparison, the resistance (not impedance) of the 70V leads of both output transformers were measured and found to be within 10% of each other. The 4, 8, and 16-ohm lines were also checked and found to be identical.

The problem shows itself with the outputs. A 1KHZ squarewave with a 10-ohm 25-watt resistor on each 8 ohm output produces only 5 volts peak to peak on the scope. It's a good, sharp square wave, but at that level, it should be. The resistors used to get warm in the good old days. Less than one-watt rms might work just fine for some headphones but is pretty lousy for speakers. The other clue is that with no signal pins 5 of the 7591s range from .2 to .67 volts with an average of only .385 volts for all 4 when there should be .8 volts at pin 5 of each.

I have replaced all of the filter capacitors, along with all components of the bias circuits and changed the chokes to 68-ohm resistors, and replaced input and inter-stage coupling capacitors and the two cathode capacitors for the 7025. I have also tested the 7591 pin resistances for shorts or opens and found none.

In short, help! It's obvious I'm missing something. I feel like I did many years ago when I was a technician for a company in Odessa, Texas many years ago. I would be going nuts working on a mini computer and my boss Woody would come back, touch a thing or two, and the thing would start working. He would note that his fingers never left his hands. Is there another Woody in the house?

I want to sell this thing I used to like on eBay to raise money for a school fund for the kids of the kids I have had in ministry for the last 14 years. All I need is $2200.00 which a 79-year-old living on Social Security just can't do.

as others have noted, have you tried lowering the AC mains voltage with a variac?
the 225 was built when typical household voltages were 110 to 117 VAC. Today many are 124VAC or higher.

 

[Mdpaudio]

The resistors in my MC225 were almost all at off value. They are carbon comp and go way off value as they age and absorb moisture. By now the tubes are worn out too.

 

[dcgillespie]

Joe -- I may be taking your comment here too literally -- but maybe not, either. When you said that you "changed the chokes to 68-ohm resistors" -- you can't just do that in the literal sense. If you did, that could certainly be a significant part if not the heart of your problem, besides the weak tubes that others have mentioned.

The earliest 225s used RF plate chokes placed in series with the 7591 plate leads to control any parasitic oscillations that might otherwise form. These chokes measure about 0.1Ω dcr each to prevent any power loss due to their presence in the plate circuit. Therefore, if you simply replaced the chokes with 68Ω resistors, that would in fact reduce your power output significantly as you are noting. It would also upset the 50/50 distribution of power developed in the plate and cathode windings of the Unity Coupled output stage.

When Mac moved away from using the chokes, the 7591 plate terminals then connect directly to their relative output transformer lead, and then the 68Ω resistors appear between the relevant output transformer lead and screen grid terminals. In other words, in the first scenario with the chokes, they are placed in series with the plate leads, while in the second scenario, the resistors are placed in series with the screen grid terminals.

As a result, if you did a simple resistor for choke change out, then that won't work, and the resistors need to be removed, and the chokes need to go back in. Or, if you want to in fact update the unit to the revised design using the resistors, I can tell you that it is a fairly involved task that requires swapping out transformer leads and even drive leads to the output tubes as I recall, if you are going to mimic the original factory wiring change that was made when the change to using resistors was made. I know this well, because I did this very work on my own 225 a few years back, and it is much more complicated that just swapping the chokes for resistors would suggest.

I would suggest then that if the chokes were simply swapped out for resistors, that is likely the heart of your problem. Or, if you did go to the considerable effort to revise the wiring so as to properly revise the unit to the updated design, then I would suggest that you go back and check that work carefully, as again, it is a rather involved process, and can easily result in problems if not done correctly.

I hope this helps!

Dave

 

[Mr Boochie]

Joe -- I may be taking your comment here too literally -- but maybe not, either. When you said that you "changed the chokes to 68-ohm resistors" -- you can't just do that in the literal sense. If you did, that could certainly be a significant part if not the heart of your problem, besides the weak tubes that others have mentioned.

The earliest 225s used RF plate chokes placed in series with the 7591 plate leads to control any parasitic oscillations that might otherwise form. These chokes measure about 0.1Ω dcr each to prevent any power loss due to their presence in the plate circuit. Therefore, if you simply replaced the chokes with 68Ω resistors, that would in fact reduce your power output significantly as you are noting. It would also upset the 50/50 distribution of power developed in the plate and cathode windings of the Unity Coupled output stage.

When Mac moved away from using the chokes, the 7591 plate terminals then connect directly to their relative output transformer lead, and then the 68Ω resistors appear between the relevant output transformer lead and screen grid terminals. In other words, in the first scenario with the chokes, they are placed in series with the plate leads, while in the second scenario, the resistors are placed in series with the screen grid terminals.

As a result, if you did a simple resistor for choke change out, then that won't work, and the resistors need to be removed, and the chokes need to go back in. Or, if you want to in fact update the unit to the revised design using the resistors, I can tell you that it is a fairly involved task that requires swapping out transformer leads and even drive leads to the output tubes as I recall, if you are going to mimic the original factory wiring change that was made when the change to using resistors was made. I know this well, because I did this very work on my own 225 a few years back, and it is much more complicated that just swapping the chokes for resistors would suggest.

I would suggest then that if the chokes were simply swapped out for resistors, that is likely the heart of your problem. Or, if you did go to the considerable effort to revise the wiring so as to properly revise the unit to the updated design, then I would suggest that you go back and check that work carefully, as again, it is a rather involved process, and can easily result in problems if not done correctly.

I hope this helps!

Dave

What is the advantage of replacing the plate chokes with 68-ohm resistors? I would think that the choke would be better for stopping certain oscillations.

Thanks!

 

[dcgillespie]

This is one of those areas where different approaches have different levels of effectiveness with different tubes in tamping down parasitic oscillations. The Gm characteristics of the tubes employed also plays heavily into the mix as well.

For example, the vast majority two tube push-pull designs using the very high Gm 8417 tube require plate stopper resistors if the design is going to remain even remotely stable. Not all, but certainly most. Even a single example of this tube when being tested in my power output tube tester requires a plate stopper for stability. On the other hand, even push-pull parallel designs using the much lower Gm 6L6 family of tubes typically requires no plate stoppers of any kind at all.

With tubes of more sane (but yet still considered high) Gm levels however (including 6550, 7591, and EL34), the more typical concern is parasitic activity (and the resulting protection) regarding the screen grid, which is most important when the screen grid operates at a DC voltage level that is very nearly that of the plate voltage. In virtually all such designs, screen stoppers are a must to prevent not only garden variety parasitic issues, but also the very destructive kind that can result in tube destroying output tube arcing.

Specific to the MC225, no doubt the original plate chokes were quite effective in eliminating/preventing parasitic oscillations in the design. My own MC225 that originally included the chokes certainly didn't display any such misbehavior. But as installed, they do nothing to prevent any parasitic activity involving (or provide any protection of) the screen grids -- which is where most such activity is prone to occur in this tube type. Since the 7591 then does not particularly show itself to require plate stoppers for stability even in push-pull parallel designs, I think Mac wisely revised the circuit to employ screen stopper resistors instead, and dispensed with the plate chokes since they would then no longer be needed. In one move, it allowed the output stage to still remain perfectly stable (mine does after the chokes were removed and the resistors installed), it provides protection for the screens that the original configuration did not/could not do, and also was no doubt more economical to boot since the resistors were likely cheaper than the chokes. Based on all my experience with this topic, I made the change in my own 225 rather quickly, and principally for the added screen grid protection the revised configuration provides.

Dave

 

[Mr Boochie]

Thanks for the nice reply!

I kinda had an idea that the cheaper resistors replacing the chokes might be part of the reason. However, I did not know about the screen grid stability factor.

 

[Joe Reeder]

Joe -- I may be taking your comment here too literally -- but maybe not, either. When you said that you "changed the chokes to 68-ohm resistors" -- you can't just do that in the literal sense. If you did, that could certainly be a significant part if not the heart of your problem, besides the weak tubes that others have mentioned.

The earliest 225s used RF plate chokes placed in series with the 7591 plate leads to control any parasitic oscillations that might otherwise form. These chokes measure about 0.1Ω dcr each to prevent any power loss due to their presence in the plate circuit. Therefore, if you simply replaced the chokes with 68Ω resistors, that would in fact reduce your power output significantly as you are noting. It would also upset the 50/50 distribution of power developed in the plate and cathode windings of the Unity Coupled output stage.

When Mac moved away from using the chokes, the 7591 plate terminals then connect directly to their relative output transformer lead, and then the 68Ω resistors appear between the relevant output transformer lead and screen grid terminals. In other words, in the first scenario with the chokes, they are placed in series with the plate leads, while in the second scenario, the resistors are placed in series with the screen grid terminals.

As a result, if you did a simple resistor for choke change out, then that won't work, and the resistors need to be removed, and the chokes need to go back in. Or, if you want to in fact update the unit to the revised design using the resistors, I can tell you that it is a fairly involved task that requires swapping out transformer leads and even drive leads to the output tubes as I recall, if you are going to mimic the original factory wiring change that was made when the change to using resistors was made. I know this well, because I did this very work on my own 225 a few years back, and it is much more complicated that just swapping the chokes for resistors would suggest.

I would suggest then that if the chokes were simply swapped out for resistors, that is likely the heart of your problem. Or, if you did go to the considerable effort to revise the wiring so as to properly revise the unit to the updated design, then I would suggest that you go back and check that work carefully, as again, it is a rather involved process, and can easily result in problems if not done correctly.

I hope this helps!

Dave

I'll go back to the chokes and see if that flies. Thank you!

 

[TriodeLuvr]

I have trouble believing that a 68Ω anode resistor would cause a significant power loss. The operating impedance is very high at this point in the circuit. Of course, if the resistor allows the tubes to oscillate, that would be a different matter. Also, are we saying the amplifier can't output more than 5V pk-pk square into 10 ohms? That's only 2.5W. The tubes would have to be in much worse condition than merely "weak" for performance to be that degraded. I suspect a measurement error, unless the amp has two blown transformers or is being operated at half voltage.

Jack

 

[TriodeLuvr]

I'll go back to the chokes and see if that flies. Thank you!

It's evident from the notes in the schematic that McIntosh had problems with stability in this model. The earliest versions only had 22mH anode chokes. Then the factory added 470Ω grid stoppers. Sometime after that, they eliminated the anode chokes and simultaneously installed 68Ω screen grid stoppers. The last configuration - stoppers at control and screen grids - is the most appropriate way to quench parasitics. My suggestion would be to update the unit to this and then troubleshoot the power problem. I would not reinstall the chokes.

Jack

 

[Mr Boochie]

It's evident from the notes in the schematic that McIntosh had problems with stability in this model. The earliest versions only had 22mH anode chokes. Then the factory added 470Ω grid stoppers. Sometime after that, they eliminated the anode chokes and simultaneously installed 68Ω screen grid stoppers. The last configuration - stoppers at control and screen grids - is the most appropriate way to quench parasitics. My suggestion would be to update the unit to this and then troubleshoot the power problem. I would not reinstall the chokes.

Jack

Here's is, I believe, that version of Mc225 schematic. (Click on image to enlarge)

 

[dcgillespie]

It's a strange error on Mac's part -- the original output stage design of the 225 was identical to that of the 240 -- except that the 225 uses much higher GM tubes which would make it much more prone to parasitic oscillations. Clearly, the plate chokes weren't enough even with the control grid stoppers. Mac was hardly unaware of these type issues however as the MC-275 -- which preceded the 225 -- used plate chokes and screen stoppers, and was stable from the get-go. Why the 225 -- with tubes of a similar Gm value as those in the 275 -- was initially released with only plate chokes to begin with then is surely a mystery. Whatever the reason, the final iteration of the output stage is clearly much more stable than any of the earlier versions.

As for the 68Ω resistors in place of the chokes, I agree that they may not be the entire reason that the problem of low power output exists, but either way, they either need to be replaced with the chokes again or better, the output stage really needs to be properly converted to the updated design if it is to meet spec again. As I say however, that is not as simple a task as it looks to be on paper. It can surely be done as I did that very exercise on my own MC-225 -- but the actual process if judged by the schematic alone is deceiving.

The OP has already stated that the B+ and bias voltages at the output stage look reasonable enough, and that the power output after the work done was on the order of 3-4 watts per channel. Since both channels started behaving in this fashion at the same time after the work was done, the chances of both OPTs going bad at the same time and in the same way are beyond inconceivable. As for the high impedance at the plates -- it is of course compared to 68Ω, but not as high as might be thought. The "effective" plate-to-plate impedance of the OPTs used in the 225 (as referred to the push-pull plates) is just 4600Ω -- much lower than most installations using 7591 tubes. That's 2300Ω per side during Class A operation, but just 1150Ω during Class B operation. Because of the lower overall reflected impedance than more traditional 7591 installations typically use, the peak plate currents are higher than normal at full power in the 225. The 68Ω resistors -- if placed in the plate circuit -- will have a greater impact in that scenario then. To have such low output however, there are very possibly other issues going on as well, and output stage oscillation may be one of them. Replacing the chokes back into the circuit then is necessary if only to rule out the potential problems of replacing them with 68Ω resistors can cause.

Dave

 

[Mr Boochie]

To have such low output however, there are very possibly other issues going on as well, and output stage oscillation may be one of them. Replacing the chokes back into the circuit then is necessary if only to rule out the potential problems of replacing them with 68Ω resistors can cause.

Dave

My bet is with parasitic oscillations in the output stage. Much of the tube's power maybe wasted in reproducing ultra-high frequency resonance.

 

[TriodeLuvr]

It's unlikely the square waves would have looked normal on the scope if the amplifier was oscillating. And to be clear, my comments about bad OPTs or operating at half power were strictly tongue-in-cheek. I was only emphasizing the unlikely possibility that the amplifier is actually limited to 2.5W output. It would be much more probable that this is a measurement error. OTOH, maybe all four screen grids have been vaporized by oscillations. Yes, that's another tongue-in-cheek hypothesis.

 

[TriodeLuvr]

Just food for thought - it wasn't all that long ago (in vintage terms) that manufacturers in general lacked the test equipment needed to detect certain defects. One example of this bit me back when I was active in ham radio. I had a commercially manufactured linear amplifier designed in the early '60s that used four 811A output tubes. It produced very severe TVI (television interference) on the VHF channels, but with 1,600W peak output, I had always assumed it was simply an overload problem. One day, for no good reason, I looked at the output of the amplifier with my 100 MHz scope. What I discovered was that a small portion of every HF sine wave had a VHF parasitic oscillation riding on it at about 70 MHz. The oscillation was short duration, and its amplitude was also small - just a little bubble riding on every cycle of the primary waveform. I eventually cleaned it up, and the TV problems cleaned up too. I'm certain this problem had existed from new, and that the manufacturer had never even been aware of it. This is one of the reasons I always say no tube audio amplifier (some of which use tubes having a bandwidth much greater than the 811A) can be known to be stable until thoroughly examined with a spectrum analyzer. Parasitic frequencies might not travel through the OPTs to any great extent, but they can certainly influence the sound and/or cause serious damage. It's only a wild guess, but perhaps this is what happened to the 225. Or perhaps Mac tested it under certain conditions, called it good, but later discovered that other conditions would encourage a problem.

Jack