Hi solderhead,
And that is it in a nutshell. New motors always spin faster than teh old ones with bad brushes and bearings.
-Chris
Not adjustable. They are set by the mag coil in relation to the 124 VDC supply. Turn it down to 118 volts or there abouts. This voltage will tend to creep up over time.
And that is it in a nutshell. New motors always spin faster than teh old ones with bad brushes and bearings.
-Chris
Just wanted to give an update on the PM-1.5 fan replacement:
The new OEM fan is definitely quieter than the original OEM fan in some respects, but louder in others, but the problem is easy enough to fix:
First, the new bearings mean no bearing noise. Quite an improvement!
Second, the new motor is made in China instead of Japan, and there are some subtle side effects of the Chinese manufacture that I had not anticipated. They make the idle speed of the fan quite a bit higher, and the fan quite a bit louder at idle:
Although the old and new motors share the exact same part number, the average DCR measurements are much higher with the new motors than with the old ones, presumably as the result of differences in wire gauge and winding methods in the motors. At the same voltage the new motors pull fewer mA than the originals, which means that they are less sensitive to the effects of the dropper resistors on the fan control board. As the amp idles the fans run at too high of a voltage and they are way to loud -- they sound like a Lockheed Constellation at takeoff!
Fixing the problem is simple. Just re-scale the voltage dividers on the fan control board to bring the operating voltage of the fan motor back into spec and the fans are nice and quiet at idle. They also speed up appropriately when the amp operates under load.
I re-scaled the dividers on my amp so that the amp with the new fan behaved like an amp that had a properly functioning original fan. The objective was to make the fan nice and quiet when the amp was at idle in the low setting (for home use), while maintaining the ability of the fan to speed-up appropriately under load. I also scaled the divider so that the fan moved a lot more air and was a lot louder under idle conditions at the high setting (for pro use), and increased speed appropriately under load. I also maintained the thermal switch that bypasses the voltage dropping resistors so that the fan will run at full voltage and full speed if the amp ever gets too warm. If and when that ever happens (I tested it by jumpering the thermal switch), the fan would move a LOT of of air, until the amp cools back down and the thermal switch opens again, allowing the fan to run at the speeds selected by the switch settings.
This way I was able to maintain adequate cooling of the amp under all conditions while keeping the noise level down to the original levels when the amp was new.
I hope this helps. If anyone needs help with this sort of fan project, drop me a line. I had to buy a box of these fans to get the right part, and now I have more than I need.
The new OEM fan is definitely quieter than the original OEM fan in some respects, but louder in others, but the problem is easy enough to fix:
First, the new bearings mean no bearing noise. Quite an improvement!
Second, the new motor is made in China instead of Japan, and there are some subtle side effects of the Chinese manufacture that I had not anticipated. They make the idle speed of the fan quite a bit higher, and the fan quite a bit louder at idle:
Although the old and new motors share the exact same part number, the average DCR measurements are much higher with the new motors than with the old ones, presumably as the result of differences in wire gauge and winding methods in the motors. At the same voltage the new motors pull fewer mA than the originals, which means that they are less sensitive to the effects of the dropper resistors on the fan control board. As the amp idles the fans run at too high of a voltage and they are way to loud -- they sound like a Lockheed Constellation at takeoff!
Fixing the problem is simple. Just re-scale the voltage dividers on the fan control board to bring the operating voltage of the fan motor back into spec and the fans are nice and quiet at idle. They also speed up appropriately when the amp operates under load.
I re-scaled the dividers on my amp so that the amp with the new fan behaved like an amp that had a properly functioning original fan. The objective was to make the fan nice and quiet when the amp was at idle in the low setting (for home use), while maintaining the ability of the fan to speed-up appropriately under load. I also scaled the divider so that the fan moved a lot more air and was a lot louder under idle conditions at the high setting (for pro use), and increased speed appropriately under load. I also maintained the thermal switch that bypasses the voltage dropping resistors so that the fan will run at full voltage and full speed if the amp ever gets too warm. If and when that ever happens (I tested it by jumpering the thermal switch), the fan would move a LOT of of air, until the amp cools back down and the thermal switch opens again, allowing the fan to run at the speeds selected by the switch settings.
This way I was able to maintain adequate cooling of the amp under all conditions while keeping the noise level down to the original levels when the amp was new.
I hope this helps. If anyone needs help with this sort of fan project, drop me a line. I had to buy a box of these fans to get the right part, and now I have more than I need.
Hi Solderhead,
That is great that you got it all figured out. Good show!
It's been a while for me (replacing fans), but your description is what we used to see, except that the fan was quieter but only a little faster. I am a little concerned that you may have a current draw somewhere that caused the fans to turn faster.
I guess I should have measured the normal voltage across the original fans. Something I didn't consider since we did have the original parts at one time. It would have helped you out.
The new motors may also be more efficient and have less friction in the bearings. That would make them turn faster also.
-Chris
That is great that you got it all figured out. Good show!
It's been a while for me (replacing fans), but your description is what we used to see, except that the fan was quieter but only a little faster. I am a little concerned that you may have a current draw somewhere that caused the fans to turn faster.
I guess I should have measured the normal voltage across the original fans. Something I didn't consider since we did have the original parts at one time. It would have helped you out.
The new motors may also be more efficient and have less friction in the bearings. That would make them turn faster also.
-Chris
anatech said:
Thanks for your help. You make a good point about the possibility of excess current draw at idle. That would cause the fans to spin much faster.
I was fortunate enough to have two amps to work on, one with a good fan and one with a bad one. I used the good amp for reference measurements of the voltages, and rescaled the fan motor's voltage divider to reach that endpoint.
The good news is that the amp is behaving well, and on the bench it puts out its rated power into a resistive load. The fan responds appropriately in testing. Although I guess there could be a current draw somewhere, I've never found it. The amp still has those old dual-element PSU caps on the low and mid rails. Those will be getting changed out very soon.
Hi solderhead,
That explains much. Replace them now, or as soon as possible. Personally I would recommend you not use the Carver until that time. Your voltage adjust may need to be redone.
Recently I worked on one where a dual cap leaked and blew. It was not a pretty sight and damage was high. Please avoid this situation.
-Chris
Wow!
That explains much. Replace them now, or as soon as possible. Personally I would recommend you not use the Carver until that time. Your voltage adjust may need to be redone.
Recently I worked on one where a dual cap leaked and blew. It was not a pretty sight and damage was high. Please avoid this situation.
-Chris
Clarification, please...
anatech,
I'm most interested in getting a more complete description of your experience with locked-up commutators; most importantly, which models/years were subject to this failure mode?
I ask for selfish reasons: I am running two M-400t amps to biamp my speakers (horizontally).
Relevant note: earlier this year I had heard vague references to the same frequency spectrum limitation you describe, with the problem being limited to "early" Carver amps. I contacted Rolland Barr of Hi-Tech Audio, Ltd (a recognized Carver repair facility) to discuss... he had no experience with this type of failure, but opined that any of the "t" models are not "early" Carver designs. He further opined that a mf/hf-only signal would be fine driving an M-400t -- he advised me to do so without fear. He was so confident that he guaranteed to repair any resulting damage gratis.
So, I'm sure both you and Rolland are WAAAYYYYYY smarter and more experienced than I will ever be on this subject, and I'm not playing any sort of "one against the other" game. I just want to get as much information as I can on this particular subject since the health of my amplifiers may be in the balance.
anatech said:
anatech,
I'm most interested in getting a more complete description of your experience with locked-up commutators; most importantly, which models/years were subject to this failure mode?
I ask for selfish reasons: I am running two M-400t amps to biamp my speakers (horizontally).
Relevant note: earlier this year I had heard vague references to the same frequency spectrum limitation you describe, with the problem being limited to "early" Carver amps. I contacted Rolland Barr of Hi-Tech Audio, Ltd (a recognized Carver repair facility) to discuss... he had no experience with this type of failure, but opined that any of the "t" models are not "early" Carver designs. He further opined that a mf/hf-only signal would be fine driving an M-400t -- he advised me to do so without fear. He was so confident that he guaranteed to repair any resulting damage gratis.
So, I'm sure both you and Rolland are WAAAYYYYYY smarter and more experienced than I will ever be on this subject, and I'm not playing any sort of "one against the other" game. I just want to get as much information as I can on this particular subject since the health of my amplifiers may be in the balance.
Hi faustus,
I'm sorry, but I have to disagree with Rolland here. Some of the later models without commutation stages - yes. That would be the Lightstar most probably. I still wouldn't do this.
The commutator timing is set to a minimum hold up time. At high midrange and high frequencies, this would tend to keep the higher rail voltages engaged (commutators locked up). What he may be thinking is that high notes may be of a level lower than the trigger voltage. Therefore he is hoping the commutators are never triggered so they can not lock. Rest assured, at high power levels running mids or above only will definitely cause thermal runaway in these amps.
Consumer users do not normally do what you are doing, not with Carver amplifiers especially. Your M-400t's are not safe running mids up at higher levels. However, you may never trigger the higher supply voltage. The answer then becomes, how loud will you get these? The circuits are all subject to this limitation.
Therefore, you may or may not run into trouble. But Rolland is incorrect in thinking these newer amps are immune to this.
-Chris
I'm sorry, but I have to disagree with Rolland here. Some of the later models without commutation stages - yes. That would be the Lightstar most probably. I still wouldn't do this.
The commutator timing is set to a minimum hold up time. At high midrange and high frequencies, this would tend to keep the higher rail voltages engaged (commutators locked up). What he may be thinking is that high notes may be of a level lower than the trigger voltage. Therefore he is hoping the commutators are never triggered so they can not lock. Rest assured, at high power levels running mids or above only will definitely cause thermal runaway in these amps.
Consumer users do not normally do what you are doing, not with Carver amplifiers especially. Your M-400t's are not safe running mids up at higher levels. However, you may never trigger the higher supply voltage. The answer then becomes, how loud will you get these? The circuits are all subject to this limitation.
Therefore, you may or may not run into trouble. But Rolland is incorrect in thinking these newer amps are immune to this.
-Chris
anatech,
Your quick reply is greatly appreciated and I will refrain from high volume blasting until I have this issue sorted-out. I sure wish I could get a consensus opinion from two or more Carver repair experts regarding the commutator-induced failures.
With the hope of engaging you further on this matter, there are a couple of other details that I can offer. First, I have heard the problem described as a "rail switching logic error," which to me would imply that the logic error could be fixed or mitigated in later designs. Have you heard it described as a logic error?
Second, most of my listening is done at moderate volume, but I do push the amps hard on at least one CD per weekend. Thus, I have already run my dual M-400t rig at very high sustained power levels, with the power level LEDs on both the LF-only amp and the MF/HF-only amp near or at the maximum (top LED). Should my MF/HF amp have failed already, or does it take cumulative sessions to induce failure?
Many thanks for any additional commentary.
Your quick reply is greatly appreciated and I will refrain from high volume blasting until I have this issue sorted-out. I sure wish I could get a consensus opinion from two or more Carver repair experts regarding the commutator-induced failures.
With the hope of engaging you further on this matter, there are a couple of other details that I can offer. First, I have heard the problem described as a "rail switching logic error," which to me would imply that the logic error could be fixed or mitigated in later designs. Have you heard it described as a logic error?
Second, most of my listening is done at moderate volume, but I do push the amps hard on at least one CD per weekend. Thus, I have already run my dual M-400t rig at very high sustained power levels, with the power level LEDs on both the LF-only amp and the MF/HF-only amp near or at the maximum (top LED). Should my MF/HF amp have failed already, or does it take cumulative sessions to induce failure?
Many thanks for any additional commentary.
Faustus,
I believe anatech is a good person to listen to regarding Carver products. No one I know knows more about them.
I believe anatech is a good person to listen to regarding Carver products. No one I know knows more about them.
Hi Cal,
Thanks. BTW, I can't think of any Jimmy Buffett song I can relate to. 😀
Hi faustus,
Could it be done today? Probably, and at greater cost. I'm not sure on this.
Luckily these normally shut down before killing the speakers. However, if the triac goes too, the amp will remain on at higher rail voltages than they were designed for. The destruction in that case would be extensive. Those mids and highs must be screaming!
If I were you, I would use a different amplifier for mid/high duty. I can't understand how your levels could be that high with normal program material in your home either.
-Chris
Thanks. BTW, I can't think of any Jimmy Buffett song I can relate to. 😀
Hi faustus,
No. It's an entirely analog circuit. The circuits will not respond fast enough for high frequencies to operate in time. Therefore the commutator must be timed to hold up for a short while to accommodate any high level high notes to avoid clipping them. So, if you continue to trigger the commutator on, it will not turn off before you trigger it again. This is not something that is even remotely easy to get around. If it was, Carver would have done so. They were on the cutting edge of this technology.
Could it be done today? Probably, and at greater cost. I'm not sure on this.
You are operating these in an area that would not be covered under warranty. Let's just say that your amps are getting hotter than they should be. At some point you may exceed the SOA on the output or driver transistors and "pop". Ex-amp.
Luckily these normally shut down before killing the speakers. However, if the triac goes too, the amp will remain on at higher rail voltages than they were designed for. The destruction in that case would be extensive. Those mids and highs must be screaming!If I were you, I would use a different amplifier for mid/high duty. I can't understand how your levels could be that high with normal program material in your home either.
The issue is thermal runaway due to higher than designed rail voltages being applied for long times across your output transistors. The commutators control the supplies switched to the output stage. I don't see how there could possibly be a disagreement here. If it makes you feel any better, I was the service manager of the Canadian Factory service center. My contacts at Carver were normally the engineers, not Carver's service department. They were not helpful at all. So, for what it's worth .....
-Chris
I guess if anyone wanted to nitpick, "switching logic" could refer to either digital switching circuits of today, or analog switching circuits of yesteryear that used transistors or even vacuum tubes.
From what I've seen in the protection circuits of Carver amps, they do use transistors in a digital logic type circuit -- the transistors are switched all-on or all-off, just as the switching would take place in a digital computer. Granted, the "computer" in a Carver amp is an extremely simple circuit that only uses a few semiconductor devices and not a high density array of transistors on a CPU, but in both cases the transistors are switched on or off in a binary fashion. Binary signaling is what defines "digital" logic, and if an analog device is biased full-on or full-off, then its behaving as a digital switch. The protection circuits in a Carver amp can be considered to be fully "digital," just as vacuum tube computers that used binary signaling were functioning in a fully digital mode, and not amplifying a signal as analog devices.
With respect to commutator timing, my understanding of the circuits is that the triggering of commutator activation was a binary event that was defined by the signal envelope either exceeding or not exceeding a pre-defined threshold. Once activated, the duration of commutator activation was based on the time constant of an RC circuit. IIRC there were some TSBs from Carver that recommended changing the values of R or C in an attempt to modify the duration of commutation activation, and that these TSBs were specifically related to the types of commutation failures that Anatech had described. Chris, does any of this sound familiar to you?
I would have to agree that there really shouldn't be any opposite opinions about commutation failures among people that understand the Carver circuits. They seem to be well known among people who have serviced the amps, or have done their due diligence in reading about them. My guess would be that the warranty repair people who dealt with the Pro Series amps in bi-amping applications would have a better knowledge of the commutator problem than someone who is rebuilding consumer gear that's already long out of warranty. The later don't have the factory engineers to use as consultants.
From what I've seen in the protection circuits of Carver amps, they do use transistors in a digital logic type circuit -- the transistors are switched all-on or all-off, just as the switching would take place in a digital computer. Granted, the "computer" in a Carver amp is an extremely simple circuit that only uses a few semiconductor devices and not a high density array of transistors on a CPU, but in both cases the transistors are switched on or off in a binary fashion. Binary signaling is what defines "digital" logic, and if an analog device is biased full-on or full-off, then its behaving as a digital switch. The protection circuits in a Carver amp can be considered to be fully "digital," just as vacuum tube computers that used binary signaling were functioning in a fully digital mode, and not amplifying a signal as analog devices.
With respect to commutator timing, my understanding of the circuits is that the triggering of commutator activation was a binary event that was defined by the signal envelope either exceeding or not exceeding a pre-defined threshold. Once activated, the duration of commutator activation was based on the time constant of an RC circuit. IIRC there were some TSBs from Carver that recommended changing the values of R or C in an attempt to modify the duration of commutation activation, and that these TSBs were specifically related to the types of commutation failures that Anatech had described. Chris, does any of this sound familiar to you?
I would have to agree that there really shouldn't be any opposite opinions about commutation failures among people that understand the Carver circuits. They seem to be well known among people who have serviced the amps, or have done their due diligence in reading about them. My guess would be that the warranty repair people who dealt with the Pro Series amps in bi-amping applications would have a better knowledge of the commutator problem than someone who is rebuilding consumer gear that's already long out of warranty. The later don't have the factory engineers to use as consultants.
I think that if someone is running into commutation failures in home environment, they'd have to be actively bi-amping (with a higher than subwoofer crossover point) into an extremely inefficient speaker system, like a ribbon. By any chance are you somebody that has Amazing Loudspeakers and tries to run them at high volume levels? That could be a recipe for disaster.
Hi solderhead,
I tried to answer the question using the most common definitions of "logic". You are correct in that it is a type of analog logic, hard wired. Early commutators were "bang-bang" controllers. Later ones tracked the output voltage, keeping the VCE voltage approximately constant on the outputs above a certain level. But they could only track up to to a certain frequency and the top tier was still a "bang-bang" controlled event. Triggering this one is where your troubles really start in mid/high use.
No change orders or service notes were seen regarding any change in RC values for commutator timing. This could have been a USA thing, but the engineers I dealt with were of the opinion that the time constants were set up for the best compromise. I agree with them. This may have occurred very early in time with the cube product. They were still learning and developing the concept. I came in a little later when the bugs were mostly worked out. I still repaired an absolute ton of cubes though. Very few people were capable of repairing them. Patience I think made the difference. Hackers simply could not succeed here.
-Chris
I tried to answer the question using the most common definitions of "logic". You are correct in that it is a type of analog logic, hard wired. Early commutators were "bang-bang" controllers. Later ones tracked the output voltage, keeping the VCE voltage approximately constant on the outputs above a certain level. But they could only track up to to a certain frequency and the top tier was still a "bang-bang" controlled event. Triggering this one is where your troubles really start in mid/high use.
No change orders or service notes were seen regarding any change in RC values for commutator timing. This could have been a USA thing, but the engineers I dealt with were of the opinion that the time constants were set up for the best compromise. I agree with them. This may have occurred very early in time with the cube product. They were still learning and developing the concept. I came in a little later when the bugs were mostly worked out. I still repaired an absolute ton of cubes though. Very few people were capable of repairing them. Patience I think made the difference. Hackers simply could not succeed here.
That was our biggest problem. Bridged no less. 🙄
For the most part - yes. Pro people could not understand that these amps did put out the advertised power. Therefore they always used the lightest amps for ..... mid/high duty. I guess they can't read, nor do they understand English. We could never win with them.
-Chris
anatech said:
That continues to be a problem to this day. IIRC the Amazings are 4R speakers, and may dip a bit lower. They're a pretty difficult load to drive.
I've heard quite a few folks claiming that its OK to bridge a moderate powered amp like an M-500t, and use it to drive the 4-ohm loads like the Amazings, just because Bob Carver did it at a trade show. Taking it all one step further, some of them are bi-amping.
IMO its hard to imagine a worse setup, and they're just asking for trouble -- Carver amps don't like low Z loads when they're running in bridged mode. When bridged they should be run at 8R and not anything less. Then you've got the commutation/thermal runaway issues if they're only driving HF signals. What kind of setup could be worse? Well, I guess they could try to do this sort of thing under the power of a small gas powered generator with a long extension cord...
Hi solderhead,
Amazings might dip to 2 or 3 ohms. They could be driven by the low impedance version of the PM 1.5, but not mid / highs only. Thinking about it, the amazing ribbon crosses over rather low, so this may be okay to do.
The problem wit ha Carver bridged into a low Z load are the demands placed on the AC supply. Over current protection would protect the amp just fine. So drooping AC is probably what would cause high current (followed by triac failure and all the fun that follows).
-Chris
The long extension cord might save the day there by limiting the current.
Amazings might dip to 2 or 3 ohms. They could be driven by the low impedance version of the PM 1.5, but not mid / highs only. Thinking about it, the amazing ribbon crosses over rather low, so this may be okay to do.
The problem wit ha Carver bridged into a low Z load are the demands placed on the AC supply. Over current protection would protect the amp just fine. So drooping AC is probably what would cause high current (followed by triac failure and all the fun that follows).
-Chris
This discussion has brought up a question- Stepping over to the Pro world for a minute, Clair Bros used Carvers for some time for biamping stage monitors and the like. (I forget the model numbers Carver made for them) Were the later, as in 10-15 years ago, Carver amps ok with mid/high duty?
Cal Weldon said:
Cal,
Thanks for the kind advice. No arguments from me on that one, anatech has convinced me to tread lightly.
solderhead said:
solderhead,
Thank you for your participation, I sure appreciate everyone's advice. Let me give a couple of details...
I am not trying any craziness with Amazings, but I am trying to biamp a pair of relatively efficient ADS L-1290 speakers (8-Ohm, 300W, 91 db) that I recently acquired.
I started by driving them with a single M-400t, but I was interested in potential sonic improvements that biamping enables. Sure enough, adding a second M-400t in a biamped config did improve the sound at all volumes -- sounds cleaner, with a more composed content and/or less extraneous content.
But I definitely do not want to damage my Cubes!
anatech,
Thank you, sincerely, for your efforts to save my M-400t from disaster. I greatly value your extensive knowledge on these designs and appreciate you taking the time to share your experience.
You're right, I'm not going crazy with the volume on a normal basis -- The "super volume" test was a one-time event to test the new system config. In fact, maybe (obviously?) I'm not pushing the MF/HF amp hard enough to induce commutator failures? The crossover frequency is 500Hz, so the majority of the work is normally being shouldered by the LF amp. If I keep the volumes within reason, I might be able to maintain the system config, as is?
Christmas is approaching, so maybe I can justify just one more amp -- after all, this is practically an emergency.
anatech, thank you, again, for the solid technical details and advice.
Thank you, sincerely, for your efforts to save my M-400t from disaster. I greatly value your extensive knowledge on these designs and appreciate you taking the time to share your experience.
anatech said:
You're right, I'm not going crazy with the volume on a normal basis -- The "super volume" test was a one-time event to test the new system config. In fact, maybe (obviously?) I'm not pushing the MF/HF amp hard enough to induce commutator failures? The crossover frequency is 500Hz, so the majority of the work is normally being shouldered by the LF amp. If I keep the volumes within reason, I might be able to maintain the system config, as is?
anatech said:
Christmas is approaching, so maybe I can justify just one more amp -- after all, this is practically an emergency.
anatech, thank you, again, for the solid technical details and advice.
carverv pro
imix, you asked about clair bros using carver amps that ran not only lp2's (double 15" cone/ single 2" compression driver) but also ran s-4 foh speaker system. amp models were pm 1.5, pm1.5a, pm 1200 and pm 2.0 clair had trouble with middle two output transistors (which are the ones that are always on) blowing up on the original pm 1,5's carver/clair came up with the solution of parrallelling an xtra pair of transistors on the middle two-this became the pm1.5a and then the pm 1200 (all the same amp-475watts/8 ohms. the 2.0 was a different animal-roughly tha same power but only 11 pounds but changed to plastic case outputs. they worked fine for low/ hi applications day in day out for over 20 years...........midas
imix, you asked about clair bros using carver amps that ran not only lp2's (double 15" cone/ single 2" compression driver) but also ran s-4 foh speaker system. amp models were pm 1.5, pm1.5a, pm 1200 and pm 2.0 clair had trouble with middle two output transistors (which are the ones that are always on) blowing up on the original pm 1,5's carver/clair came up with the solution of parrallelling an xtra pair of transistors on the middle two-this became the pm1.5a and then the pm 1200 (all the same amp-475watts/8 ohms. the 2.0 was a different animal-roughly tha same power but only 11 pounds but changed to plastic case outputs. they worked fine for low/ hi applications day in day out for over 20 years...........midas