anatech said:
How about this one - the "better amp" is the one that can be turned up higher without distorting....
Dunno how many people I've heard that one from... !!
Cheers
Hi Clem,
Didn't know you were watching. 😉
-Chris
Didn't know you were watching. 😉
Know we know why most of the stuff sold is unlistenable.
-Chris
Hmmm, bad news on my part guys, i just finnished replacing the resistors on my display from the pm 1.5, now the right channel leds are really acting up, they seem more accurate indeed, the 500ms attack and 50ms attack are really going at it...so i'm starting to beleive the problem is coming somewhere from the power supply board, i'm checking the schematics are as i'm writing this, i'll be following the traces to see what makes the right channel light up faster, i tried a 100hz wave at a peak of about 300w, man i love the smell of heated voice coils, there's about 5.40vrms difference between both channels, i seem to have a larger difference between channels as i go higher in power...anyways, i'll post my results shortly...
Thanks, any suggestions are appreciated,
By the way, i would'nt call myself a display freaks as i have little indactors on my pre-amps mixers and other devices, my bgw's have only power and clipping indicators, but why not use them if they are there ? I understand they can't all be like the ones used in Mc'intosh amps where behind the analog needles there's an actual digital circuit that compensates for the impedance load, signal noise and clipping factor...but i still love my carver amps...
😎
Thanks, any suggestions are appreciated,
By the way, i would'nt call myself a display freaks as i have little indactors on my pre-amps mixers and other devices, my bgw's have only power and clipping indicators, but why not use them if they are there ? I understand they can't all be like the ones used in Mc'intosh amps where behind the analog needles there's an actual digital circuit that compensates for the impedance load, signal noise and clipping factor...but i still love my carver amps...
😎
Hi pjaneiro,
The scaling resistors are on the power supply pcb. Make sure all your solder connections are good as well.
-Chris
The scaling resistors are on the power supply pcb. Make sure all your solder connections are good as well.
-Chris
Huh? Is this with a "Y"'d input connection, or the preamp in mono?? And is this the voltage difference at the loads? It sounds to that before this channel-to-channel voltage difference did not exist. I believe this was when the inputs were "Y" connected. Now, there is a 5.4 Vrms voltage difference between channels? .....You've got to have a poorly tracking volume control.
How is this test being carried out? Do you have a CD with tone bursts?? ...capturing on a scope, or is this a setting on the Carver??
Hi mrshow4u,
Yes, something has gone badly in that amp. Could have been a bad connection with the probe even.
The time constants for the meters are very roughly set by fixed components. Always remember that the meters are there to give you an idea of the levels. They never were intended to be a measuring insterment.
-Chris
Yes, something has gone badly in that amp. Could have been a bad connection with the probe even.
The time constants for the meters are very roughly set by fixed components. Always remember that the meters are there to give you an idea of the levels. They never were intended to be a measuring insterment.
-Chris
My tests are run using a signal generator using a y connector at the amp, i noticed that at low levels i have little difference between both channels, but when i ran the test with my 18" woofers at high levels just around 300w there is a 5v difference, but display indicates the right channel is higher even at low levels...
I'll check for the scaling resistors on my pw pcb when i come back from work...
I'll check for the scaling resistors on my pw pcb when i come back from work...
Hi pjaneiro,
Interesting measurements.
Do you have dummy loads you can try? You may get some results at lower powers with a 4 ohm load. I would also suggest looking at the output with a scope and search for distortion. You may have some emitter resistors high in value or another fault, or over current protection may be interferring (that circuit can be defective).
-Chris
Interesting measurements.
Do you have dummy loads you can try? You may get some results at lower powers with a 4 ohm load. I would also suggest looking at the output with a scope and search for distortion. You may have some emitter resistors high in value or another fault, or over current protection may be interferring (that circuit can be defective).
-Chris
Okay, the "5.40 Vrms" isn't such a big deal when your pumping out 300 Watts worth of voltage swing. The meters are reading voltage and may b calibrated for watts at one load only. I'm assuming it's calibration is for 8 ohms. 300 Watts into 8 ohms gives a voltage swing of 49 Vrms. If your channel to channels error is 5.4 Vrms, that is about a dB. Not too bad. That falls in line with your channel to channel mismatch with a lower reference level having a 80 mV difference. So, I think there is a 1 dB or so gain difference between amp channels. 1dB is not a large difference. You will get larger differences when acoustic gain is applied to the left and right speakers. Most rooms don't apply an even amount of acoustic gain to the left and right speakers. The acoustic gain will vary with frequncy too.
You could spend a lot of time to super-match your meter and your amp gain. That certainly would reduce the psychological contribution of the percieved imbalance, but acoustically it's always there and very complicated.
Most of these tests can be done at low to moderate volumes. ..there's no need to smell voice coil varnish.
You could spend a lot of time to super-match your meter and your amp gain. That certainly would reduce the psychological contribution of the percieved imbalance, but acoustically it's always there and very complicated.
Most of these tests can be done at low to moderate volumes. ..there's no need to smell voice coil varnish.
That's right, my right channel was running at about 55vrms and my left channel was at about 50vrms or so, i know 1db is'nt much, but it's enough to make me go nuts when those lights are playing, most of the time i am running this amp at about 100wrms, and at that level i barely have a 1.5v diff, this may seem stupid for some but i'm about ready to take the right channel board out and re-furb it to brand new....i noticed one thing when i measure the mv on resistor r87 and r49 on my right channel it runs at about 30mv but on my left channel they are at about 15mv,
I think the quickest way to get the matching better would be to replace the following components with tight tolerance varieties.
1. The input coupling cap, if one exists.
2. the feed back resistor.
3. The feedback shunt resistor and feedback shunt cap.
I don't have the schematic, so I can't call out component designations.
Think of the amp having a bunch of open loop gain and the feedback components will set the gain. So, a component tolerance problem here could give you your gain error (...albiet small)
1. The input coupling cap, if one exists.
2. the feed back resistor.
3. The feedback shunt resistor and feedback shunt cap.
I don't have the schematic, so I can't call out component designations.
Think of the amp having a bunch of open loop gain and the feedback components will set the gain. So, a component tolerance problem here could give you your gain error (...albiet small)
Hi pjaneiro,
Hi mrshow4u,
Pjaneiro, check the gain setting resistors while you are at it. THe one coming back from the output may be changing it's resistance as the voltage across it goes up, or the 470 uF (C4) cap is flaky.
-Chris
See this post. and fix that. No load and no signal, remember? Short the input jacks to make sure.
Hi mrshow4u,
No, but it's out of spec. They should be within 1/4 of a dB at the worst.
Pjaneiro, check the gain setting resistors while you are at it. THe one coming back from the output may be changing it's resistance as the voltage across it goes up, or the 470 uF (C4) cap is flaky.
-Chris
Hi mrshow4u,
Just saw your post. Yes, that's where I'm going with this.
R58, R3 and C4 would set your gain. The input coupling cap is a film type (C35 0.33 uF), but the solder joints may crack on this part. It should be fastened to the amp board with silicon or something similar.
R58 (130 K) should have a larger wattage than the original installed. Figure it sees close to 250V p-p when you are really cranked up. In the Lightstar, this resistor used to burn out. We are talking about a hole in the board. I believe most of those were caught. I did all the Canadian units I knew about.
The feedback "shunt resistor" (R3 910 ohms) is critical. It's more important that you match two together for each channel than the actual value (although that helps keep it in spec with the other amps).
-Chris
Just saw your post. Yes, that's where I'm going with this.
R58, R3 and C4 would set your gain. The input coupling cap is a film type (C35 0.33 uF), but the solder joints may crack on this part. It should be fastened to the amp board with silicon or something similar.
R58 (130 K) should have a larger wattage than the original installed. Figure it sees close to 250V p-p when you are really cranked up. In the Lightstar, this resistor used to burn out. We are talking about a hole in the board. I believe most of those were caught. I did all the Canadian units I knew about.
The feedback "shunt resistor" (R3 910 ohms) is critical. It's more important that you match two together for each channel than the actual value (although that helps keep it in spec with the other amps).
-Chris
for the resistor part that's what i had done at that time, when i check the mv across the emiter transistors they are at 3.4mv, but if i take a value from the resistor itself at r89 or r47, it gives me 30mv compared to 15mv on the left channel, i do not know what those resistors are for, i've read in the schematics manual they check for high freqency shutdown....
For the moment i am not able to verify this hands on, but as soon as i get back home, i'll chekc it out...
For the moment i am not able to verify this hands on, but as soon as i get back home, i'll chekc it out...
when you repair units , do you leave any marks ? maybe this amp passed in your hands in the past ? 😉 this amps has been worked on , and from what i can see a lot of people....
unlike most people i know i prefer to ask than just throw something at it...
unlike most people i know i prefer to ask than just throw something at it...
Hi pjaneiro,
The resistors you are referring to do not make sense to me. You may be measuring something that is not central to what we are doing. I read R47 as a 680 R resistor.
When I service an amplifier you usually can not tell I was ever there. It still looks factory unless the damage made some burn marks. I consider that a trait of a good service guy. With other amplifiers I will often put the differential transistors together and use a little heatsink grease. Not applicable to these.
So I guess my mark, is no mark. 😉
PM me with your serial number and I can check against the units I did service. I am sure the owner's name was different back then.
-Chris
The resistors you are referring to do not make sense to me. You may be measuring something that is not central to what we are doing. I read R47 as a 680 R resistor.
When I service an amplifier you usually can not tell I was ever there. It still looks factory unless the damage made some burn marks. I consider that a trait of a good service guy. With other amplifiers I will often put the differential transistors together and use a little heatsink grease. Not applicable to these.
So I guess my mark, is no mark. 😉
PM me with your serial number and I can check against the units I did service. I am sure the owner's name was different back then.
-Chris
The resistors i am talking about are the ceramic ones r49 and r87
if you measure the top end of the resistor with chassis you will have a mv value, on my right channel i am getting 30mv and on my left channel 15mv, in my first posts i misread the manual and was mesuring the resistors themselves instead of the emiters on the transistors, even after adjusting the bias i still have that difference...
btw, i double checked in my manual, and nowhere do i find the r58 of 130k i need to check, actually i have no resistor of 130k at all on my amp board...
if you measure the top end of the resistor with chassis you will have a mv value, on my right channel i am getting 30mv and on my left channel 15mv, in my first posts i misread the manual and was mesuring the resistors themselves instead of the emiters on the transistors, even after adjusting the bias i still have that difference...
btw, i double checked in my manual, and nowhere do i find the r58 of 130k i need to check, actually i have no resistor of 130k at all on my amp board...
Whoa, whoa, whoa, whoa.
Make sure you are just measuring across the resistor. Do not measure from ground to one end of the resistor. You are interested in reading the current through a resistor, not the voltage from ground. DC offsets will spoil any bias readings referenced from ground.
When measuring bias:
Make sure the unit has warmed up and stabilized (20 mins)
disconnect the load. This will remove any DC current path, due to offset, into the load that will corrupt the measurement.
If you have paralleled output transistors, take an average of all of the emmiter currents. Even with degeneration resistors ("Emmiter Resistors") and Vbe matching, the output transistors will not have even emmiter currents. I'm not sure in the case of the 1.5... ....it might have a single pair of output devices with a few commutators.
Adjust bias
Keep an eye on the set for the next 5 minutes.
Re-check bias. If ok, check after 20 mins.
Don't think of your bias control as having anything to do with volume. It can have an impact on volume, but let's not go there, I don't think that extreme under-biasing is your problem.
After the bias is properly set, forget about it.
Bias is one of those adjustments where a little goes a long way. A poor adjustment here and (as marketing would say) you've just entitled yourself to the opportunity to install a handful of fresh parts.
If you want to tweak the 1 dB gain error out start with the R58, R3, and C4. I don't have the service manual, but I think Chris was stating that these are the feedback components. Think of those components as a "gain wrapper" around a properly working amp. I think your amp is working fine, but the components that set the gain may be a little out of spec. There is too much working correctly on the amp for me to think that there is anything wrong in the amp itself. ....it sounds like negative feedback problems. six components for two channels. That shouldn't be so bad. Worth a shot.
Make sure you are just measuring across the resistor. Do not measure from ground to one end of the resistor. You are interested in reading the current through a resistor, not the voltage from ground. DC offsets will spoil any bias readings referenced from ground.
When measuring bias:
Make sure the unit has warmed up and stabilized (20 mins)
disconnect the load. This will remove any DC current path, due to offset, into the load that will corrupt the measurement.
If you have paralleled output transistors, take an average of all of the emmiter currents. Even with degeneration resistors ("Emmiter Resistors") and Vbe matching, the output transistors will not have even emmiter currents. I'm not sure in the case of the 1.5... ....it might have a single pair of output devices with a few commutators.
Adjust bias
Keep an eye on the set for the next 5 minutes.
Re-check bias. If ok, check after 20 mins.
Don't think of your bias control as having anything to do with volume. It can have an impact on volume, but let's not go there, I don't think that extreme under-biasing is your problem.
After the bias is properly set, forget about it.
Bias is one of those adjustments where a little goes a long way. A poor adjustment here and (as marketing would say) you've just entitled yourself to the opportunity to install a handful of fresh parts.
If you want to tweak the 1 dB gain error out start with the R58, R3, and C4. I don't have the service manual, but I think Chris was stating that these are the feedback components. Think of those components as a "gain wrapper" around a properly working amp. I think your amp is working fine, but the components that set the gain may be a little out of spec. There is too much working correctly on the amp for me to think that there is anything wrong in the amp itself. ....it sounds like negative feedback problems. six components for two channels. That shouldn't be so bad. Worth a shot.
The bias is fine on both channels now, i was just wondering why on my right channel i would have different values when i measured different points on the board compared to the other board, i will be replacing the parts for the input section on both channels tonight so i have the same parts on the two channels...
Hi mrshow4u,
The one thing I have to pound into a tech's head is that bias and offset may drift after an adjustment. Most are in too much of a rush to set these things properly. Mind you, the labour charge is a little higher to do things right. Not a bad trade in the long run (to do it right).
Hi pjaneiro,
Bad contacts with your probes can easily throw off you measurements. I would expect them to be low. When measuring, select a stable test point so your probes don't slip and blow the amp up. I've seen many repaired amps die on the bench because the tech slipped setting bias. Normally the final touch up because he isn't paying as much attention. Costly error. If you can't find stable spots, attach clip leads or solder connections with the amp off, power up and test. Having many meters helps here. Amps like Adcoms need this treatment.
-Chris
Yep.
The one thing I have to pound into a tech's head is that bias and offset may drift after an adjustment. Most are in too much of a rush to set these things properly. Mind you, the labour charge is a little higher to do things right. Not a bad trade in the long run (to do it right).
Hi pjaneiro,
Bad contacts with your probes can easily throw off you measurements. I would expect them to be low. When measuring, select a stable test point so your probes don't slip and blow the amp up. I've seen many repaired amps die on the bench because the tech slipped setting bias. Normally the final touch up because he isn't paying as much attention. Costly error. If you can't find stable spots, attach clip leads or solder connections with the amp off, power up and test. Having many meters helps here. Amps like Adcoms need this treatment.
-Chris
- Status
- Not open for further replies.