anatech said:
Actually I was making the assumption that the amp provided proper control for the fan. That was probably a foolish assumption on my part- the fan control is probably no more reliable than the rest of the amp. If you think impure thoughts about it, it will probably die a fiery death.
I_F
Hi I_Forgot,
I needed that today, thanks!
The output from the mag coil depends on the flux density (or current through it). As the amplifier delivers power, the primary current increases (a lot). All windings connected to the core get more energy, or their voltage rises. Voltage feedback is returned to the error amp from all rails of the amplifier section so that they are fairly stable.
Now, the fans run on their own little winding. If you were to increase the voltage enough, there would be a point where the fans all run slowly. Naturally this is too high for the rails. The fans have almost enough voltage to turn under normal idle conditions. As the amplifier works into a load you can hear the fan(s) speeding up and slowing down to the music. You have to hear a wall of these doing this at a big show. It's pretty funny!
So, what's the problem you might ask? Easy. Most DC fans have a speed control circuit inside. This means at low voltage the fan supply is too lightly loaded. I don't know if that matters, except for higher voltages and under load you may over volt the filter caps, the current draw being lower that the originals. If you did away with the rectifiers and used the AC from the mag coil, those fans would tend to want to lock to the supply frequency. I think they would act as a normal transformer and overheat with the chopped wave form.
So, you are basically into uncharted waters there. The amp simply controls the fan with varying voltage as the load increases.
-Chris
I needed that today, thanks!
The output from the mag coil depends on the flux density (or current through it). As the amplifier delivers power, the primary current increases (a lot). All windings connected to the core get more energy, or their voltage rises. Voltage feedback is returned to the error amp from all rails of the amplifier section so that they are fairly stable.
Now, the fans run on their own little winding. If you were to increase the voltage enough, there would be a point where the fans all run slowly. Naturally this is too high for the rails. The fans have almost enough voltage to turn under normal idle conditions. As the amplifier works into a load you can hear the fan(s) speeding up and slowing down to the music.
You have to hear a wall of these doing this at a big show. It's pretty funny!So, what's the problem you might ask? Easy. Most DC fans have a speed control circuit inside. This means at low voltage the fan supply is too lightly loaded. I don't know if that matters, except for higher voltages and under load you may over volt the filter caps, the current draw being lower that the originals. If you did away with the rectifiers and used the AC from the mag coil, those fans would tend to want to lock to the supply frequency. I think they would act as a normal transformer and overheat with the chopped wave form.
So, you are basically into uncharted waters there. The amp simply controls the fan with varying voltage as the load increases.
-Chris
SleeperSupra said:
The SC PCR series of amps are great. a little of the mosfet mist on the top end but very warm. They use the good Hitachi Mosfets. 2Sj50/K135's which i think were the best sounding of that series.
The Carvers can be harsh and very annoying. we had them in our studio and i could mark spots on the master fader where the sound would be harsh. presumably from the rail switching in the amp.
The SC are great little amps. the beat out the Adcoms hands down. they are a killer super budget amp.
The MA series and anything with a 4 digit number A2502 etc etc used some form of class H. and are not in the same league as the PCR or PM series or the Pro Power series of amps. all use the exact same circuit.
The Little SC's are still one of my fav amps of all time. I have even accidentally shorted out the speaker cables with out killing the amp. just popped a fuse but the amp still worked.
Hi Zero Cool,
Please PM me. No parts have showed up yet. 2nd request.
-Chris
Yes, totally different circuit to the Carver product.
The early ones - for sure! The late model ones were a completely different animal and sounded very good. So it depends on what you are comparing.
Again, different series and different models had their own sound with Adcom. The statement was too sweeping so as to be inaccurate for all models.
Please PM me. No parts have showed up yet. 2nd request.
-Chris
Banned
Joined 2002
What about the Carver Cubes were they good ? they were slick and cute,
Like this one..http://cgi.ebay.com/CARVER-CUBE-M-4...ryZ39783QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
Were they ticking time bombs tho ?
I also like this model,
http://cgi.ebay.com/Carver-TFM-55x-...ryZ39783QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I think the amp that i played with when i djid was a PWM model, i dunno tho. I do know that QSC uses PWM now tho.
Like this one..http://cgi.ebay.com/CARVER-CUBE-M-4...ryZ39783QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
Were they ticking time bombs tho ?
I also like this model,
http://cgi.ebay.com/Carver-TFM-55x-...ryZ39783QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
I think the amp that i played with when i djid was a PWM model, i dunno tho. I do know that QSC uses PWM now tho.
Mine has been working fine since about 1982. I sent it is once for service to have the triac firing adjusted. It was buzzing quite a bit. But that's about it.
If you don't mistreat it in any of the way previously discussed (dumb human tricks), but just use it like a normal home audio product you should be fine.
If you don't mistreat it in any of the way previously discussed (dumb human tricks), but just use it like a normal home audio product you should be fine.
\
I think it is interesting that Doug Self addresses this in his Class-G article and design. He says it results from diode switching noise. He claims to greatly reduce it by using Schottkys. Sounds like a potential fix assuming the rail switching in the unit you mention uses accomplished by diodes (it is in the M-400). But is there a Schottky that is safe to use at the voltages involved?
Hi Guys,
Carver Cubes (M-400 and variants) are very tricky to service. Most people break the darn sockets. I hate when they do that! Too much grease is another pet peeve I have. Keep these away from most technicians.
The triac firing (uneven) has been an issue with those. The triac had been upgraded. They were the worst sounding models, but they did do exactly what they were designed to do.
Hi sam9,
What I can say is this. The "glitches" on the output are caused by the commutating transistors. Playing with diodes is not the answer.
I am so sick of people "solving" problems with diodes!!!!
What happens is that the next voltage is applied to the output stage as a stepped waveform in the earlier models. What this means is when the signal crosses a threshold, BANG, the voltage is switched suddenly to the next level. This causes a glitch to be transmitted through the output transistor (the step is normally around 30 VDC to 50 ~ 60 VDC) to the output line. Later units then went to a tracking supply configuration which eliminated this glitch to a great extent. The final tier was still a "bang-bang" type. No diodes were played with because they were not the problem to begin with. It was how the supply voltage increases where handled.
So now you know what the problem was. I had countless amps on my bench and investigated this issue on my own. They had it fixed before any comment was made by myself.
Pretty smart cookies over at Carver eh? <------(I'm Canadian, what can I say?)
-Chris
Carver Cubes (M-400 and variants) are very tricky to service. Most people break the darn sockets. I hate when they do that! Too much grease is another pet peeve I have. Keep these away from most technicians.
The triac firing (uneven) has been an issue with those. The triac had been upgraded. They were the worst sounding models, but they did do exactly what they were designed to do.
Hi sam9,
No. I wish people would study things more completely before offering an opinion on them. I have his book, but haven't read it yet.
What I can say is this. The "glitches" on the output are caused by the commutating transistors. Playing with diodes is not the answer.
I am so sick of people "solving" problems with diodes!!!!
What happens is that the next voltage is applied to the output stage as a stepped waveform in the earlier models. What this means is when the signal crosses a threshold, BANG, the voltage is switched suddenly to the next level. This causes a glitch to be transmitted through the output transistor (the step is normally around 30 VDC to 50 ~ 60 VDC) to the output line. Later units then went to a tracking supply configuration which eliminated this glitch to a great extent. The final tier was still a "bang-bang" type. No diodes were played with because they were not the problem to begin with. It was how the supply voltage increases where handled.
So now you know what the problem was. I had countless amps on my bench and investigated this issue on my own. They had it fixed before any comment was made by myself.
Pretty smart cookies over at Carver eh? <------(I'm Canadian, what can I say?)
-Chris
Since you haven't read the book yet, the section of interest is Chapter 10. In the text around figures 10.7 and 10.8 diode gliches are discussed and the two figures are Spice simulations showing them. They are different from the normal crossover glitches. Sadly, he did not show traces from an actual amplifier before and after using Schottky's.
I felt I was offering a speculation rather than an opinion. Sorry if I wasn't clear on that point. Until such time as I get the nerve to muck around in side the thing I am speculating. If I were ever to try it and report on the results, then it would be an opinion.
Hi sam9,
No, I didn't take your statement as your opinion. You clearly stated that you read this in Self's book. I'm okay with that.
What I did was to offer my direct observations and the corporate response from Carver to those problems. So what I stated was exactly what I measured on my bench with my 'scope and THD analyzer. I cued up the glitches with the tracking supply commutation. Newer amplifiers were the commutated supply tracked the audio signal in an analog fashion still had these glitches, but to a greatly lesser degree.
Please feel free to make comments like this. I was unaware of Mr. Self's comments and so sought to set the record straight. I do appreciate this type of input.
I will have a peek later. One thing is correct though, these do not look like normal crossover glitches. Mostly because they aren't.
-Chris
No, I didn't take your statement as your opinion. You clearly stated that you read this in Self's book. I'm okay with that.
What I did was to offer my direct observations and the corporate response from Carver to those problems. So what I stated was exactly what I measured on my bench with my 'scope and THD analyzer. I cued up the glitches with the tracking supply commutation. Newer amplifiers were the commutated supply tracked the audio signal in an analog fashion still had these glitches, but to a greatly lesser degree.
Please feel free to make comments like this. I was unaware of Mr. Self's comments and so sought to set the record straight. I do appreciate this type of input.
I will have a peek later. One thing is correct though, these do not look like normal crossover glitches. Mostly because they aren't.
-Chris
Self's Class G schematic is visually simpler than an M400 and he doesn't say anything about power supplies so I'm guessing they are conventional with voltages taken off the secondary after the appropriate number of windings.
He also does something clever with biasing. Both inner and outer drivers are biased from the outer rail. Apparently biasing the4 inner driver from the inner rail is another source of switching noise. (I don't have the book nearby so my recollection is uncertain.) This hardly looks like a "simple mod" even if it could be don at all.
He also does something clever with biasing. Both inner and outer drivers are biased from the outer rail. Apparently biasing the4 inner driver from the inner rail is another source of switching noise. (I don't have the book nearby so my recollection is uncertain.) This hardly looks like a "simple mod" even if it could be don at all.
Hi sam9,
I glanced over the example used in Self's book. It's not the same as the Carver. Just related.
In Carver amps, the drivers are run from the top rails to begin with, I'm not sure about the M-400 though. The step voltage is also much higher than is normally used in a Carver. Perhaps that is why there are three rails as opposed to two.
So he is examining a much earlier Hitachi design rather than a Carver design, the differences between the two are important. I'll read that section more carefully once I begin this book.
-Chris
I glanced over the example used in Self's book. It's not the same as the Carver. Just related.
In Carver amps, the drivers are run from the top rails to begin with, I'm not sure about the M-400 though. The step voltage is also much higher than is normally used in a Carver. Perhaps that is why there are three rails as opposed to two.
So he is examining a much earlier Hitachi design rather than a Carver design, the differences between the two are important. I'll read that section more carefully once I begin this book.
-Chris
I love my Carver amps!
I'm a Business and Electrical Engineering student at Lehigh University, and I have been acquiring broken Carver TFM amps over the past couple of years for repair.
I had been using my first TFM-15 amp for deejaying, although it just recently developed an issue (see http://www.diyaudio.com/forums/showthread.php?s=&threadid=106952). I just recently repaired a TFM-15CB, which will be used for redundancy with my deejay equipment (repair guide here: http://www.lehigh.edu/~evb209/projects/carver2/carver2.html) I also have two broken TFM-25's, which I have been trying to combine into one working unit. I'm sure DIYaudio.com will be a great for these and other projects!
I'm a Business and Electrical Engineering student at Lehigh University, and I have been acquiring broken Carver TFM amps over the past couple of years for repair.
I had been using my first TFM-15 amp for deejaying, although it just recently developed an issue (see http://www.diyaudio.com/forums/showthread.php?s=&threadid=106952). I just recently repaired a TFM-15CB, which will be used for redundancy with my deejay equipment (repair guide here: http://www.lehigh.edu/~evb209/projects/carver2/carver2.html) I also have two broken TFM-25's, which I have been trying to combine into one working unit. I'm sure DIYaudio.com will be a great for these and other projects!
back to the subject of fans:
Chris, I finally obtained some replacement fans for the PM-1.5 from the OEM. I had the problem of spent bearings on an older fan, which had developed an intermittent noise problem. Sometimes the fan would idle slowly and quietly, and at other times it would squeal horribly. I cleaned a lot of dirt out of the fan, but the bearings were shot and it needed to be replaced.
After a lot of searching I found the OEM fan and ordered a batch of them. Upon replacing the original fan, the new fan is quite a bit louder at idle. Compared to the original 20-year old fans which would spin very slowly at idle votlages of 6-9 VDC, the new fan spins a LOT faster and it moves a LOT more air at and idle votlage of 18 VDC. Notice that when the amp is idle the DC voltage across the fan is a lot higher than it was on the original fan that was replaced.
I have to admit I'm a bit surprised by the fact that the supply voltage increased so dramatically upon replacing the fan. It jumped from somewhere between 6 - 9 VDC with the old fan to about 18 VDC with the new fan. As you would expect, the new fan spins a lot faster and moves a lot more air. But it seems far too loud, both at idle and under load.
By any chance do you know the votlage specifications for the fan supply? It would be interesting to know the voltage specifications for operation at load vs. no-load condtions, at both the Low and Hi switch settings. It would also be interesting to know the voltage specs for the fan when the thermal switch bypasses the dropping resistors on the fan board. FWIW the fan is rated 12-32 VDC / 80mA, though at idle the old fans were markedly under voltage at 6-9 VDC. To get to such low voltages with the new fan one would need to sub the dropping resistors, and at those votlages, sometimes the new fan will turn and sometimes it won't
anatech said:The output from the mag coil depends on the flux density (or current through it). As the amplifier delivers power, the primary current increases (a lot). All windings connected to the core get more energy, or their voltage rises. Voltage feedback is returned to the error amp from all rails of the amplifier section so that they are fairly stable.
Now, the fans run on their own little winding. If you were to increase the voltage enough, there would be a point where the fans all run slowly. Naturally this is too high for the rails. The fans have almost enough voltage to turn under normal idle conditions. As the amplifier works into a load you can hear the fan(s) speeding up and slowing down to the music.
So, what's the problem you might ask? Easy. Most DC fans have a speed control circuit inside. This means at low voltage the fan supply is too lightly loaded. I don't know if that matters, except for higher voltages and under load you may over volt the filter caps, the current draw being lower that the originals. If you did away with the rectifiers and used the AC from the mag coil, those fans would tend to want to lock to the supply frequency. I think they would act as a normal transformer and overheat with the chopped wave form.
So, you are basically into uncharted waters there. The amp simply controls the fan with varying voltage as the load increases.
Chris, I finally obtained some replacement fans for the PM-1.5 from the OEM. I had the problem of spent bearings on an older fan, which had developed an intermittent noise problem. Sometimes the fan would idle slowly and quietly, and at other times it would squeal horribly. I cleaned a lot of dirt out of the fan, but the bearings were shot and it needed to be replaced.
After a lot of searching I found the OEM fan and ordered a batch of them. Upon replacing the original fan, the new fan is quite a bit louder at idle. Compared to the original 20-year old fans which would spin very slowly at idle votlages of 6-9 VDC, the new fan spins a LOT faster and it moves a LOT more air at and idle votlage of 18 VDC. Notice that when the amp is idle the DC voltage across the fan is a lot higher than it was on the original fan that was replaced.
I have to admit I'm a bit surprised by the fact that the supply voltage increased so dramatically upon replacing the fan. It jumped from somewhere between 6 - 9 VDC with the old fan to about 18 VDC with the new fan. As you would expect, the new fan spins a lot faster and moves a lot more air. But it seems far too loud, both at idle and under load.
By any chance do you know the votlage specifications for the fan supply? It would be interesting to know the voltage specifications for operation at load vs. no-load condtions, at both the Low and Hi switch settings. It would also be interesting to know the voltage specs for the fan when the thermal switch bypasses the dropping resistors on the fan board. FWIW the fan is rated 12-32 VDC / 80mA, though at idle the old fans were markedly under voltage at 6-9 VDC. To get to such low voltages with the new fan one would need to sub the dropping resistors, and at those votlages, sometimes the new fan will turn and sometimes it won't
Hi solderhead,
The voltage across the fans is determined by both the loading on the fan supply and the load the amplifier is under. They normally turn slowly at idle. Your B+ adjustment will affect how the fans run. Therefore, your B+ adjustment must be set before you can do anything else. That includes bias current. I guess you can set up the meters though.
Do you have a link for the OEM supplier?
Hi crazygreek,
-Chris
The voltage across the fans is determined by both the loading on the fan supply and the load the amplifier is under. They normally turn slowly at idle. Your B+ adjustment will affect how the fans run. Therefore, your B+ adjustment must be set before you can do anything else. That includes bias current. I guess you can set up the meters though.
Do you have a link for the OEM supplier?
Hi crazygreek,
I strongly disagree with this practice. What was wrong with the other amp?
-Chris
AFAIK the B+ is only adjustable on the 124 VDC rail by adjusting RP1 on the PSU board. It is adjusted properly to 124VDC, and the OTs are properly biased.
I seem to be missing how to adjust the B+ on the low (36VDC) and medium (74VDC) rails. If you could point that out I'd appreciate it.
The fan circuit supply is actually taken off of the lowest voltage taps on the PT (red wires that supply the 36 VDC rail). There's a set of diodes forming a full wave SS rectifier on the back side of the PSU board, and the rectified unfiltered AC leaves the main PSU board at point C. It travels along a red wire under the chassis to the fan board. No adjustment of the fan supply voltage is available on either the main PSU board or the fan board. The only way to adjust the fan B+ is by swapping out the voltage dropping resistors on the fan PCB.
I seem to be missing how to adjust the B+ on the low (36VDC) and medium (74VDC) rails. If you could point that out I'd appreciate it.
The fan circuit supply is actually taken off of the lowest voltage taps on the PT (red wires that supply the 36 VDC rail). There's a set of diodes forming a full wave SS rectifier on the back side of the PSU board, and the rectified unfiltered AC leaves the main PSU board at point C. It travels along a red wire under the chassis to the fan board. No adjustment of the fan supply voltage is available on either the main PSU board or the fan board. The only way to adjust the fan B+ is by swapping out the voltage dropping resistors on the fan PCB.