Further DH200 mods
There are a few components that can be changed for a smoother sound quality,
I suggest for the I/P capacitor a BLACK GATE 10mfd 50volt non polarized type
with a 0.47mfd across it.. across the O/P coil a CADDACK resistor,non inductive type
in place of r36.. replace C8 across the amplified diode fdrom 0.1 to 2.0mfds
I suggest you use the small layer block caps of 1mfd 50volt types, one will fit the board directly the other can be tacked onto the back quite easily, the black gate, and caddack are available from MICHEAL PERCY USA.As has already been suggested
and crucial is to change the mains cable to something more substantial 2.5-4mm^2
Change the mains I/P fuse to a 5amp TIMED type. The SLO BLOW types can have
a very large internal resistance caused by internal helixs of wire to slow down the
heating effect of typical USA brown out surges, but are sound quality killers.
Next to consider changing the O/P speaker fuse for a simple relay controlled safety
circuit, one that comes to mind is from RANDY SLONE'S book HIGH-POWER AUDIO
CONSTRUCTION MANUAL PAGE 282. the circuit shown can be simplified if you don't
want the condition status, just remove the astable multivibrator, it really is a simple
circuit, so get the book from your local library or buy it, it has much to offer even
to those smarter than the average bear,in my experience a major failure of the
amplifier board will often put the full supply voltage across your speaker, usually the
fuse survives but the speaker is disemboweled even a 2amp fuse can survive,
any fuse is a sound killer but a necessary evil so the relay option is worth considering. Make sure the relay has beefy multiple contacts to cut contact resistance and the sound will be unaffected. More mods next time regards
There are a few components that can be changed for a smoother sound quality,
I suggest for the I/P capacitor a BLACK GATE 10mfd 50volt non polarized type
with a 0.47mfd across it.. across the O/P coil a CADDACK resistor,non inductive type
in place of r36.. replace C8 across the amplified diode fdrom 0.1 to 2.0mfds
I suggest you use the small layer block caps of 1mfd 50volt types, one will fit the board directly the other can be tacked onto the back quite easily, the black gate, and caddack are available from MICHEAL PERCY USA.As has already been suggested
and crucial is to change the mains cable to something more substantial 2.5-4mm^2
Change the mains I/P fuse to a 5amp TIMED type. The SLO BLOW types can have
a very large internal resistance caused by internal helixs of wire to slow down the
heating effect of typical USA brown out surges, but are sound quality killers.
Next to consider changing the O/P speaker fuse for a simple relay controlled safety
circuit, one that comes to mind is from RANDY SLONE'S book HIGH-POWER AUDIO
CONSTRUCTION MANUAL PAGE 282. the circuit shown can be simplified if you don't
want the condition status, just remove the astable multivibrator, it really is a simple
circuit, so get the book from your local library or buy it, it has much to offer even
to those smarter than the average bear,in my experience a major failure of the
amplifier board will often put the full supply voltage across your speaker, usually the
fuse survives but the speaker is disemboweled even a 2amp fuse can survive,
any fuse is a sound killer but a necessary evil so the relay option is worth considering. Make sure the relay has beefy multiple contacts to cut contact resistance and the sound will be unaffected. More mods next time regards
Sorry for being "late to market" on this. In a "back to back," a Belles Series One (BSO) spanked a 200 in bass response.
In a later "B2B" between a BSO and a 280, bass was near identical w/the BSO having a very small edge. The flip side: 280's mids & highs had (or rather had less of) an edge.
I own 2 (each) BSOs and 280s. BSOs for my subs; the 280s to biamp my L & R channels in my theatre. Looking to acquire a third 280 for the middle channel when its finished.
when I get some time I'd like to do most of the simpler (read that lower co$t) mods I've read in this & other forums.
Just my (crazy) $.02
Final (hopefully) update on my voltage regulator for PC-19C board in Hafler 220.
Lots of post on this, but this one is for a repeat of the listening test as I describe in post #339 and some comments.
After the last test, I realized that one the regulators (LM337) had somehow shorted and the rail voltage went up to about 85. I suspect this high rail voltage caused the fuses to blow during the first test. I replaced the LM337, replaced the potentiometers with matched zeners (68 volt), and re-installed the VR, hooking up only the right channel. After normal use for several weeks, I checked the rail voltage again. and then again after several hours of loud use (while working outside). No problems. Then I repeated the listening test as described in post #339, page 34. Same outcome, although the hammer sound sounded different (better, clearer, less harsh) to me. The difference was hardly night and day, but noticeable. No fuses blew.
I have since hooked up the right channel, and am using the amp as usual.
If anyone wants to build a voltage regulator, using the 1993 Audio Amateur article as a guide, I would make the following changes:
The holes for R1, R2, D3, D4 coud be spread out so that the devices lay flat. I think the regulators may have shorted to one of the leads of these devices via the clamp-on heat sink while handling the board.
As suggested on this forum (jackinnj), I replaced D1 and D2 with 33 volt zeners. This seemed to solve the voltage surge at turn-on that ruined one of my PC 19 boards.
There is no tolerance to shorting of the outputs. So far (many weeks) this does not seem to matter, but be careful during installation and testing. The Maida circuit (see post #288, page 29), has resistors to protect against shorts.
I combined the two channels onto one board that sits on top of a toroid transformer. I don't know how the boards would fit with the stock transformer; the original AA article had them in an external box. The transformers on the VR get a little warm, but not enough for me to be concerned about. The PC19 board draws about 30 mA per rail, and the transformer is rated at 100mA; there is one transformer for each channel. Someone suggested using one transformer for both channels, but I think the one transformer (Triad FP120-100) rated at 12VA would be under-powered and would get too warm. Also, the main benefit of a VR would be negated if too small a transformer is used, as voltage could sag too much during big peaks.
So now it seems that I have a reliable voltage regulator that makes a little difference in sound. DH200's or DH220's with a stock power supply may benefit more than mine which has a larger transformer and twice the capacitance on the main power supply. Thanks to all for your suggestions and help.
Lots of post on this, but this one is for a repeat of the listening test as I describe in post #339 and some comments.
After the last test, I realized that one the regulators (LM337) had somehow shorted and the rail voltage went up to about 85. I suspect this high rail voltage caused the fuses to blow during the first test. I replaced the LM337, replaced the potentiometers with matched zeners (68 volt), and re-installed the VR, hooking up only the right channel. After normal use for several weeks, I checked the rail voltage again. and then again after several hours of loud use (while working outside). No problems. Then I repeated the listening test as described in post #339, page 34. Same outcome, although the hammer sound sounded different (better, clearer, less harsh) to me. The difference was hardly night and day, but noticeable. No fuses blew.
I have since hooked up the right channel, and am using the amp as usual.
If anyone wants to build a voltage regulator, using the 1993 Audio Amateur article as a guide, I would make the following changes:
The holes for R1, R2, D3, D4 coud be spread out so that the devices lay flat. I think the regulators may have shorted to one of the leads of these devices via the clamp-on heat sink while handling the board.
As suggested on this forum (jackinnj), I replaced D1 and D2 with 33 volt zeners. This seemed to solve the voltage surge at turn-on that ruined one of my PC 19 boards.
There is no tolerance to shorting of the outputs. So far (many weeks) this does not seem to matter, but be careful during installation and testing. The Maida circuit (see post #288, page 29), has resistors to protect against shorts.
I combined the two channels onto one board that sits on top of a toroid transformer. I don't know how the boards would fit with the stock transformer; the original AA article had them in an external box. The transformers on the VR get a little warm, but not enough for me to be concerned about. The PC19 board draws about 30 mA per rail, and the transformer is rated at 100mA; there is one transformer for each channel. Someone suggested using one transformer for both channels, but I think the one transformer (Triad FP120-100) rated at 12VA would be under-powered and would get too warm. Also, the main benefit of a VR would be negated if too small a transformer is used, as voltage could sag too much during big peaks.
So now it seems that I have a reliable voltage regulator that makes a little difference in sound. DH200's or DH220's with a stock power supply may benefit more than mine which has a larger transformer and twice the capacitance on the main power supply. Thanks to all for your suggestions and help.
gp4jesus
Hi gp4, the older hafler200 was quite different to the later 280, the latter, if memory
serves had three pairs of O/P devices compared to the two pairs for the 200, it also
boasted more power had a larger power supply etc.. not quite fair to compare in this way..later i will be discussing ways to make the 200 give your 280 a run for its money.I would also point out that the high frequency could really only be judged with
speakers that had an extented treble response, our testing was done on three different marques of transmission line speakers all of which had selected Super-tweeters taking the response up to 30+khz and bass down to 15hz,these
were all my design, two types were of conventional folded BAILEY LINES, the third
a triangulated line which gave a much better resultant bass.The mods i am suggesting cannot be used for the DH280 other than for component improvements all of which have been tried and tested.
Hope you don't mind me refering to MMERIG on your response...
Just a thought about your regulated supply, it is ABSOLUTLY ESSENTIAL to have
reversed diodes across these regulators, so that on any condition no more than the forward bias voltage of 0.6volt can exist across the regulator if you are using the regulator outside of its normal voltage range, this is because the rate of decay of charge from the O/P side of the regulater can be slower than the I/P side causing
breakdown, of course if they are in place check their polarity and remember that
small signal diodes 1n4148 not good enough too low voltage at 75volts
Hope this of some use regards...
Hi gp4, the older hafler200 was quite different to the later 280, the latter, if memory
serves had three pairs of O/P devices compared to the two pairs for the 200, it also
boasted more power had a larger power supply etc.. not quite fair to compare in this way..later i will be discussing ways to make the 200 give your 280 a run for its money.I would also point out that the high frequency could really only be judged with
speakers that had an extented treble response, our testing was done on three different marques of transmission line speakers all of which had selected Super-tweeters taking the response up to 30+khz and bass down to 15hz,these
were all my design, two types were of conventional folded BAILEY LINES, the third
a triangulated line which gave a much better resultant bass.The mods i am suggesting cannot be used for the DH280 other than for component improvements all of which have been tried and tested.
Hope you don't mind me refering to MMERIG on your response...
Just a thought about your regulated supply, it is ABSOLUTLY ESSENTIAL to have
reversed diodes across these regulators, so that on any condition no more than the forward bias voltage of 0.6volt can exist across the regulator if you are using the regulator outside of its normal voltage range, this is because the rate of decay of charge from the O/P side of the regulater can be slower than the I/P side causing
breakdown, of course if they are in place check their polarity and remember that
small signal diodes 1n4148 not good enough too low voltage at 75volts
Hope this of some use regards...
Originaly posted by humble:
Thanks for the comments, humble, but I am confused by them.
What do you mean by "ABSOLUTELY ESSENTIAL"? There are at least two published schematics showing zener diodes across the LM337 or LM317 regulators, and these are to limit the voltage across the regulator. I don't see how an ordinary diode would limit voltage unless as you say, it is placed backwards, but then it seems there would be very little range of adjustment (0.6 volts). Why would one use the regulator outside its voltage normal range?
Originally I had 1n4007 across the regulator (input to output), and I still have these across the output and adjustment pins. The Maida paper (National Semiconductor Linear Brief #47) schematic shows 1n4001 diodes (max. forward voltage = 50) across these latter pins. This seems to be adequate as the maximum difference in voltage is 37, but I think the 1n4003 (200 v rating) is safer. Faichild data sheets show 100 volts maximum for 1n4148, not 75.
Thanks for the comments, humble, but I am confused by them.
What do you mean by "ABSOLUTELY ESSENTIAL"? There are at least two published schematics showing zener diodes across the LM337 or LM317 regulators, and these are to limit the voltage across the regulator. I don't see how an ordinary diode would limit voltage unless as you say, it is placed backwards, but then it seems there would be very little range of adjustment (0.6 volts). Why would one use the regulator outside its voltage normal range?
Originally I had 1n4007 across the regulator (input to output), and I still have these across the output and adjustment pins. The Maida paper (National Semiconductor Linear Brief #47) schematic shows 1n4001 diodes (max. forward voltage = 50) across these latter pins. This seems to be adequate as the maximum difference in voltage is 37, but I think the 1n4003 (200 v rating) is safer. Faichild data sheets show 100 volts maximum for 1n4148, not 75.
Hi mmerig,
humble is referring to good engineering practice. The zener diodes you saw in the app. note will conduct when they are reversed biased (the input-output voltage differential is too high). They will also conduct in a similar way to a regular diode. This would be the case if your primary DC supply decays faster than your regulated supply. It prevents your regulator from being reverse biased (and destroyed).
Most of us will design this same fucntion using regular diodes. They should not ever get into reverse breakdown in normal operation. Some people use 1N4001, others a little higher. I use 1N4007 only because there is hardly much of a price difference between the two and I don't feel like buying more than one type in the same family.
So,
-Chris
humble is referring to good engineering practice. The zener diodes you saw in the app. note will conduct when they are reversed biased (the input-output voltage differential is too high). They will also conduct in a similar way to a regular diode. This would be the case if your primary DC supply decays faster than your regulated supply. It prevents your regulator from being reverse biased (and destroyed).
Most of us will design this same fucntion using regular diodes. They should not ever get into reverse breakdown in normal operation. Some people use 1N4001, others a little higher. I use 1N4007 only because there is hardly much of a price difference between the two and I don't feel like buying more than one type in the same family.
So,
Good engineering practice and cheap insurance. In other words, make sure you do this. The zeners will work fine as long as they don't conduct in normal operation.
-Chris
Thanks Chris, your explanation makes more sense, as I was not experienced enough to understand humble's comments.
So, if I want to retain the advantage of the 33 volt zener across the input and output pins, and want to prevent reverse biasing the regulator, can I put an 1n4003 in series with the zener, and between the O/I pins? The ordinary diode will add about 0.7 volts, but that little change is okay with me.
As for the primary DC supply decaying faster then the regulated supply:
Do you mean the primary supply as being the big-current supply from the original transformer, bridge rectifer, and 10K uf caps?
This big current supply now goes to the MOSFETs only and seems isolated from the front end board that runs off the regulated supply. Given this, under what circumstances would the regulators become reverse biased?
So, if I want to retain the advantage of the 33 volt zener across the input and output pins, and want to prevent reverse biasing the regulator, can I put an 1n4003 in series with the zener, and between the O/I pins? The ordinary diode will add about 0.7 volts, but that little change is okay with me.
As for the primary DC supply decaying faster then the regulated supply:
Do you mean the primary supply as being the big-current supply from the original transformer, bridge rectifer, and 10K uf caps?
This big current supply now goes to the MOSFETs only and seems isolated from the front end board that runs off the regulated supply. Given this, under what circumstances would the regulators become reverse biased?
Hi mmerig,
-Chris
If I understand you correctly, no need. Your zener diode will conduct the same way a regular diode will. Don't worry.
Yes, correct. That's if it supplies the front end also.
Easy, a diode shorts or a cap shorts. In years ahead your filter caps may degrade to a point where they don't hold a charge. Anything that will cause a charge to be retained on your regulated side longer than the primary side will force reverse breakdown. Not pretty when it happens.
-Chris
HaflerDH200mods cont..
The next step suggested, is to consider improving the gain of the hafler circuit, I recommend this not be undertaken unless you feel competent to understand all the
detail I shall give.These will be in 3 stages.The one failure of the hafler circuit is the use of rather slow drivers and predriver transistors..2N5550,2N3440 devices
[FT15mhz] and limited power dissipation at 1Watt, the TO5 can makes it difficult to get big enough heat sinks to go further than hafler did which was to drive them at 10M/A for the predrivers and 8M/A for the drivers, these limits cannot safely be exceeded.Added gain will increase feedback reduce distortion and improve damping factor.[1]R12 and R14,560ohms in the CCS feeding the I/P differential pairs can be increased to 620ohms this will reduce the current through Q1 and Q5 to about 0.9M/A dropping the voltage across their
collector loads R6,R17 from about 2.2volts to 2volts, this will in turn reduce the current through the predrivers Q8,Q11 to about 9M/A, which will now enable us to reduce the values of R27,R33, from 100ohms to 90ohms[A 1kohm resistor across
the 100ohms will give 90.9ohms] this will increase gain by about 10%,we will have kept the same ratio of current with no further heat dissipation through the predrivers
Stage [2] As [1] but more involved. Change all the collector resistors of all the I/P
diff. pairs to 2Kohms from 2.2kohms[R6,R7,R17,R19] remember to reduce the phase advance resistors,R8,R18, to 300ohms from 330ohms to maintain same phase relationship this will drop the voltage on diff collectors by a further 0.2volts,thus as before enabling us to reduce current through predrivers, now we can drop R27,R33 to 82ohms giving a further increase in gain to about 15%.
The voltage across R27,R33,should be about 0.82volts thus maintaining 10M/A
through predrivers, as before...Stage 3 follows next episode
The next step suggested, is to consider improving the gain of the hafler circuit, I recommend this not be undertaken unless you feel competent to understand all the
detail I shall give.These will be in 3 stages.The one failure of the hafler circuit is the use of rather slow drivers and predriver transistors..2N5550,2N3440 devices
[FT15mhz] and limited power dissipation at 1Watt, the TO5 can makes it difficult to get big enough heat sinks to go further than hafler did which was to drive them at 10M/A for the predrivers and 8M/A for the drivers, these limits cannot safely be exceeded.Added gain will increase feedback reduce distortion and improve damping factor.[1]R12 and R14,560ohms in the CCS feeding the I/P differential pairs can be increased to 620ohms this will reduce the current through Q1 and Q5 to about 0.9M/A dropping the voltage across their
collector loads R6,R17 from about 2.2volts to 2volts, this will in turn reduce the current through the predrivers Q8,Q11 to about 9M/A, which will now enable us to reduce the values of R27,R33, from 100ohms to 90ohms[A 1kohm resistor across
the 100ohms will give 90.9ohms] this will increase gain by about 10%,we will have kept the same ratio of current with no further heat dissipation through the predrivers
Stage [2] As [1] but more involved. Change all the collector resistors of all the I/P
diff. pairs to 2Kohms from 2.2kohms[R6,R7,R17,R19] remember to reduce the phase advance resistors,R8,R18, to 300ohms from 330ohms to maintain same phase relationship this will drop the voltage on diff collectors by a further 0.2volts,thus as before enabling us to reduce current through predrivers, now we can drop R27,R33 to 82ohms giving a further increase in gain to about 15%.
The voltage across R27,R33,should be about 0.82volts thus maintaining 10M/A
through predrivers, as before...Stage 3 follows next episode
HaflerDH200 mods cont..
In stage 3 we will consider improving the LINEARITY of the predriver stage to improve the thermal stability and give better thermal tracking but still with an overal
increase in gain by about 12%.All the mods to the I/P stage will need to be done as before, however a few changes to the predrivers will be necessary...The emitter resistors of Q8,Q11,the predrivers 2N5550,2N3440,that is R27,R33 will need to be split into two resistors, from each emitter take a 75ohm resistor and in series with it
place a small resistor of 5.6ohms and then take this back to the supply rail,now to this junction with the 75ohms and the 5.6 ohms take back the 1.8kohm R26 feeding the emitter of Q7, and do the same of course for the other rail Q10, and R32, this will have the effect of Q7 and Q10
drawing part of their current through the 5.6 ohm resistor as well as the current supplied to Q11 and Q8 causing the two transistors to regulate each other acting
like a ccs, this will improve the linearity of the darlington pairs, again with no significant increase in overal dissipation of the heatsinks.
Here is a postscript also to some of my earlier comments, when I suggested using a 0.22ohm resistor soldered across a fuse to set current, of course the fuse needs to have blown. the TIMED fuse I suggested in place of SLO BLOW types was for the U.K
market for the U.S.A. a 10AMP type needs to be used.The black gate I/P capacitor
i suggested can be the small NX type which is a lower voltage range, small but good
sonically. more mods later regards
In stage 3 we will consider improving the LINEARITY of the predriver stage to improve the thermal stability and give better thermal tracking but still with an overal
increase in gain by about 12%.All the mods to the I/P stage will need to be done as before, however a few changes to the predrivers will be necessary...The emitter resistors of Q8,Q11,the predrivers 2N5550,2N3440,that is R27,R33 will need to be split into two resistors, from each emitter take a 75ohm resistor and in series with it
place a small resistor of 5.6ohms and then take this back to the supply rail,now to this junction with the 75ohms and the 5.6 ohms take back the 1.8kohm R26 feeding the emitter of Q7, and do the same of course for the other rail Q10, and R32, this will have the effect of Q7 and Q10
drawing part of their current through the 5.6 ohm resistor as well as the current supplied to Q11 and Q8 causing the two transistors to regulate each other acting
like a ccs, this will improve the linearity of the darlington pairs, again with no significant increase in overal dissipation of the heatsinks.
Here is a postscript also to some of my earlier comments, when I suggested using a 0.22ohm resistor soldered across a fuse to set current, of course the fuse needs to have blown. the TIMED fuse I suggested in place of SLO BLOW types was for the U.K
market for the U.S.A. a 10AMP type needs to be used.The black gate I/P capacitor
i suggested can be the small NX type which is a lower voltage range, small but good
sonically. more mods later regards
Hi djk,
The Leach has a different sound, but the Hafler sounds good as well. Depends on your tastes. You could do a whole lot worse than the Hafler you know.
Fix the Hafler and if you want, build up a Leach to try in that chassis. Then you aren't rushed and you can compare the two on the cheap.
-Chris
Long live Darwin!!!
The Leach has a different sound, but the Hafler sounds good as well. Depends on your tastes. You could do a whole lot worse than the Hafler you know.
Fix the Hafler and if you want, build up a Leach to try in that chassis. Then you aren't rushed and you can compare the two on the cheap.
-Chris
Hafler response
Hi gp4jesus, you are correct only in one sense about all the changes to the hafler,
It was a comment I originally made Implying that it no longer resembled a hafler.
However all the changes I recommend is still using the original circuit, but if you like
I am simply tidying up. remember the original circuit borrowed from ERNO BORBELY,
and LENDER was around in 1979, this was a significant departure from other
circuits in that it was a fully symetrical design using the rather new MOSFETS,the overal gain from its circuit topography was relatively low compared to its Hitachi
counterpart at the time but nevertheless gave a better performance in many ways.
This circuit used today using much faster drivers and predrivers TO126 of 10watt
rating and FT of 150mhz with improved heatsinking can give an outstanding performance and implementing the mods i have suggested makes it difficult to
better for the overal cost, that is why I have adopted this circuit with 4 pairs of mosfets and a stabilized high volage supply in my monoblocks of 300 watts.
I do have another integrated amp of 100 watts per chan which i and a colleague
designed in 1997, for general use using BJT'S, and despite using SANKEN 40+mhz
O/P devices I still cannot get the TOP that the Mosfets give,I play a lot of records
and with my TL speakers with supertweeters the treble is simply astonishing makes
compact disc sound very limited{in my opinion, so no debate please}.
anyway there are a few more mods to come yet, but still using the same circuit,
but more to do with the power supplies.
regards
Hi gp4jesus, you are correct only in one sense about all the changes to the hafler,
It was a comment I originally made Implying that it no longer resembled a hafler.
However all the changes I recommend is still using the original circuit, but if you like
I am simply tidying up. remember the original circuit borrowed from ERNO BORBELY,
and LENDER was around in 1979, this was a significant departure from other
circuits in that it was a fully symetrical design using the rather new MOSFETS,the overal gain from its circuit topography was relatively low compared to its Hitachi
counterpart at the time but nevertheless gave a better performance in many ways.
This circuit used today using much faster drivers and predrivers TO126 of 10watt
rating and FT of 150mhz with improved heatsinking can give an outstanding performance and implementing the mods i have suggested makes it difficult to
better for the overal cost, that is why I have adopted this circuit with 4 pairs of mosfets and a stabilized high volage supply in my monoblocks of 300 watts.
I do have another integrated amp of 100 watts per chan which i and a colleague
designed in 1997, for general use using BJT'S, and despite using SANKEN 40+mhz
O/P devices I still cannot get the TOP that the Mosfets give,I play a lot of records
and with my TL speakers with supertweeters the treble is simply astonishing makes
compact disc sound very limited{in my opinion, so no debate please}.
anyway there are a few more mods to come yet, but still using the same circuit,
but more to do with the power supplies.
regards
Would you guys give me a few pointers on the repair of a DH-200?
I have posted many times on this thread and done mods and minimal repairs, checked voltages, etc. But, now I have a DH-200 with one channel that does not play very loudly. How do I know which parts to check for lack of gain?
I have checked wiring, output MOSFETs, voltages at critical spots, etc. But, without a signal generator and o'scope I fear I have no way of fixing it. I can't just throw parts at it until I stumble on a solution.
Also, my soldering station is minimal and I've not had much luck in removing transistors for checking with the meter.
What might be the cause of almost no gain?
Thanks.......
I have posted many times on this thread and done mods and minimal repairs, checked voltages, etc. But, now I have a DH-200 with one channel that does not play very loudly. How do I know which parts to check for lack of gain?
I have checked wiring, output MOSFETs, voltages at critical spots, etc. But, without a signal generator and o'scope I fear I have no way of fixing it. I can't just throw parts at it until I stumble on a solution.
Also, my soldering station is minimal and I've not had much luck in removing transistors for checking with the meter.
What might be the cause of almost no gain?
Thanks.......
Electrolytic cap in the feedback loop to ground is open, amp goes to unity gain (but sounds fine).
C5 (470uF) on a DH200.
http://www.hafler.com/techsupport/pdf/DH-200_amp_man.pdf
C5 (470uF) on a DH200.
http://www.hafler.com/techsupport/pdf/DH-200_amp_man.pdf
djk
Thanks for the hint about the feedback cap (C5). I replaced it with a recent pull from a working circuit. No change.
This channel sounds like one that has a blown speaker fuse, only higher frequencies at a very low volume level get through. So, I jumpered over the speaker protection fuse leg of the circuit (eyelets 8 and 11). No change.
Took some time to reflow many solder joints on the rear of the circuit board, especially those around the driver transistors. All voltages to gate/drain of MOSFETs are OK.
The circuit does have -60 mV DC offset. Idling current (bias) adjustment seemed to go OK and it is set at 250 mV.
Did spot checks on voltages expected at various circuit points and they were OK. The voltage table in the DH-220 is close to the voltages supposed to be present at comparable spots in the DH-200 circuit so it is a useful reference.
Hold the presses!
Just replaced C1, the tiny electrolytic used as an input DC block cap. Installed a pull from the parts bin that came from a working amp.
Success. Now that channel sounds a tad louder than the other but perhaps some time should be spent playing the amp to get caps to settle down.
How "bad" is a DC offset of 60 mV? Considering the hassle and expense of getting matched pairs for Q1/2 and Q5/6 I am tempted to leave it alone. I don't notice speaker cone movement when switching the amp on or off and the speaker protection fuses don't pop. Am I foolish to just ignore it for now?
Thanks for your attention to my long ramble.
Dick
Thanks for the hint about the feedback cap (C5). I replaced it with a recent pull from a working circuit. No change.
This channel sounds like one that has a blown speaker fuse, only higher frequencies at a very low volume level get through. So, I jumpered over the speaker protection fuse leg of the circuit (eyelets 8 and 11). No change.
Took some time to reflow many solder joints on the rear of the circuit board, especially those around the driver transistors. All voltages to gate/drain of MOSFETs are OK.
The circuit does have -60 mV DC offset. Idling current (bias) adjustment seemed to go OK and it is set at 250 mV.
Did spot checks on voltages expected at various circuit points and they were OK. The voltage table in the DH-220 is close to the voltages supposed to be present at comparable spots in the DH-200 circuit so it is a useful reference.
Hold the presses!
Just replaced C1, the tiny electrolytic used as an input DC block cap. Installed a pull from the parts bin that came from a working amp.
Success. Now that channel sounds a tad louder than the other but perhaps some time should be spent playing the amp to get caps to settle down.
How "bad" is a DC offset of 60 mV? Considering the hassle and expense of getting matched pairs for Q1/2 and Q5/6 I am tempted to leave it alone. I don't notice speaker cone movement when switching the amp on or off and the speaker protection fuses don't pop. Am I foolish to just ignore it for now?
Thanks for your attention to my long ramble.
Dick
Hi Dick,
I always match those diff pairs. They are a pain, but when the transistors are all matched (that includes the complimentary diff pair) the amp both sounds better and has lower DC offset and drift.
I would say it's worthwhile for sure.
Why are you using "pulls" for your caps? They are not expensive to buy.
-Chris
I always match those diff pairs. They are a pain, but when the transistors are all matched (that includes the complimentary diff pair) the amp both sounds better and has lower DC offset and drift.
I would say it's worthwhile for sure.
Why are you using "pulls" for your caps? They are not expensive to buy.
-Chris
anatech,
I have some matched sets for the diff amp and was in a hurry to test whether certain caps were bad so used 'pulls.'
This is not my amp. I am rejuvenating it for a guy who found it in a dumpster. I will check with him as to how much money he wants to spend on it. But, I am sure that a lot of these old amps are being used with 60 mV of DC offset.
Whatta' hobby!
Dick
I have some matched sets for the diff amp and was in a hurry to test whether certain caps were bad so used 'pulls.'
This is not my amp. I am rejuvenating it for a guy who found it in a dumpster. I will check with him as to how much money he wants to spend on it. But, I am sure that a lot of these old amps are being used with 60 mV of DC offset.
Whatta' hobby!
Dick
Hi Dick,
I'll bet many are sitting higher than 100 mV.
And lastly .......
Just some observations. I'm sure you are well aware of this, so I probably only irritated you. Not my intent Dick. I was just surprised by your questions and some statements.
-Chris
So am I.
I'll bet many are sitting higher than 100 mV.Fair enough. I wonder if he knows how lucky he is?
Normally my first step.
Well, you are charging to repair it. Business?
And lastly .......
I normally just replace them as I've already condemned them. The vast majority I see are bad. If you look at your time spent, it would have been cheaper to simply swap them out for new ones.
Just some observations. I'm sure you are well aware of this, so I probably only irritated you. Not my intent Dick. I was just surprised by your questions and some statements.
-Chris