hi - got the audiohead off the shelf at the weekend... definately the worst amp to work on ever 
i had repaired the burned tracks, and replaced all the burned resistors.
all the small transistors checked out ok but i replaced the error amp and the LM311's as they cost only a few pence.
i also replaced the 3 preset pots for the same reason.
there a too really iritating problems with testing the amp... firstly, the confined space, but the most annoying is that due to the omision of a seperate 15V supply, you cannot easily run the amp up off a variac.
each power card has 3 or 4 low voltage supplies, all derrived by using dropper resitor + zener from the ht rails
this means you have to change the dropper reisitors for each voltage range - most irritating.
got the thing running up to 80V at the transformer primary, so decided to try full mains voltage (ok, i got bored of changing resistors).
It worked for a minute or two, then the 25A fuses blew (surprisingly gently) and testing showed the loss of 1 +ve, and 4 -ve o/p devs.
i didn't get time to examine the bias current or whether the rail switches were correct. but maybe i'll have another go next weekend.
awfull amp !
i had repaired the burned tracks, and replaced all the burned resistors.
all the small transistors checked out ok but i replaced the error amp and the LM311's as they cost only a few pence.
i also replaced the 3 preset pots for the same reason.
there a too really iritating problems with testing the amp... firstly, the confined space, but the most annoying is that due to the omision of a seperate 15V supply, you cannot easily run the amp up off a variac.
each power card has 3 or 4 low voltage supplies, all derrived by using dropper resitor + zener from the ht rails
this means you have to change the dropper reisitors for each voltage range - most irritating.
got the thing running up to 80V at the transformer primary, so decided to try full mains voltage (ok, i got bored of changing resistors).
It worked for a minute or two, then the 25A fuses blew (surprisingly gently) and testing showed the loss of 1 +ve, and 4 -ve o/p devs.
i didn't get time to examine the bias current or whether the rail switches were correct. but maybe i'll have another go next weekend.
awfull amp !
Yes the ah2400 I have here has the fuses soldered into the fuseholders!
- due to the small capacity of said holders, as per your post above.
Haven't had a chance to look at the 1400 since the new year as I've had this viral laryngitis and have felt awful.
Perhaps by next weekend. A bit of hope now I have something to work with as regards diagrams. cheers.
- due to the small capacity of said holders, as per your post above.Haven't had a chance to look at the 1400 since the new year as I've had this viral laryngitis and have felt awful.
Perhaps by next weekend. A bit of hope now I have something to work with as regards diagrams. cheers.
Incorrect email address
Hi Guys,
I'm sorry to report that I mis-typed my email address in my original post, so I had no idea anyone had been trying to contact me.
It should have been grahamprie@aol.com
Hope this helps, Graham.
Hi Guys,
I'm sorry to report that I mis-typed my email address in my original post, so I had no idea anyone had been trying to contact me.
It should have been grahamprie@aol.com
Hope this helps, Graham.
Hi all
I'm new to the forum but like what i'm seeing/reading so far.
I currently have an Audiohead AH2400 with one channel down, I can see 2 resistors in my amp that sit side by side that appear to have burn marks on and around them, i'm assuming this is the problem but need to find out why they have blown, i'm a novice at electronics and most of this thread lost me LOL. The resistors sit on the board next to the 2 output drivers (I think thats what they are) that drive the output trannies.
I'm looking for a circuit diagram or information as to which/what other amp is the same as this one in order to find a diagram.
Regards
Nelix
I'm new to the forum but like what i'm seeing/reading so far.
I currently have an Audiohead AH2400 with one channel down, I can see 2 resistors in my amp that sit side by side that appear to have burn marks on and around them, i'm assuming this is the problem but need to find out why they have blown, i'm a novice at electronics and most of this thread lost me LOL. The resistors sit on the board next to the 2 output drivers (I think thats what they are) that drive the output trannies.
I'm looking for a circuit diagram or information as to which/what other amp is the same as this one in order to find a diagram.
Regards
Nelix
Hi - i had a very big fight with an audiohead AH4000 and lost.
I'm afriad these amps aren't designed very well.
they are based on a circuit used by several different manufacturers including behringer and QSC.
i'm afraid this isn't a good amp to cut your teeth on.
From experience, the burned resistors aren't the cause of the fault, they are the result of several output devices failing short circuit.
your first job would be to identify the topology of the amp..
this 2400 is known as a class-H (or more correctly in this case, a Class-G - but they are quite similar).
The amp has a balanced input buffer (IC) then the sound goes to the level pot, in into the error amp (another op-amp IC).
The inputs of the error amp are:
the sound you want (from the level pot) and...
the sound you're getting (from the output of the amp, reduced by approx 26dB by the feedback resistor)
the 2 signals are subtracted from each other, and the resulting error signal is used to drive the output stage.
the error signal is also used to drive the clip led (to show too much error).
this is pretty much the same for most amps.
on this amp, the error goes into the bases of a pair of driver transistors which are in series across the high-voltage power rails, between them is the bias circuit (a few dides in series, some of which are in thermal contact with the heatsink for stability) which stops crossover distortion by making sure the opposing sets of output divices are always slightly on (with a quiescent current flowing through them).
there is a preset (~100R) there wich sets the bias current - easy to spot because it has a thermistor across it. (on the heatsink too).
There are 2 other presets ~1K - each with a diode in series with them - these limit the drive, thus limiting the output current to the speaker.
the drivers will be T0-220 devices (probably plastic 2SA1943 / 2SC4793 or metal tab MJ15023/33) these are the opposite way round to what you'd expect - so the N-Type (+ve rail) output are driven by a P-Type predriver.
depending on whether audiohead have modified the circuit from the qsc or not, there may be a second stage of driver transistors between the first pair and the output devices.
What makes this a class-g amp is that there are 2 sets of high voltage rails.
the output devices are fed from +/- 55V, but when its needed the rails are switched up to +/-110V - just for the part of the waveform that needs it.
this is done by a pair of FET transistors acting as switches whith their drive circuits sat monitoring the output stage drive voltage.
this reduces thermal wastage in the output stage and is rather clever.
the other clever feature of this range of amps is that the amp actually drives the power supply, not the speaker. difficult to explain, but it means that all the output devices have their collectors at ground potential which means they can be fixed to the heatsink without insulation.
This is great, but it means that the amp is a pig to test (the ground reference is actually the LS+ terminal and you need a floating scope (not reccomended for novices).
In the AH4000 i was looking at, AH had decided to ditch the grounded heatsink so that they could add an earth lift switch... so all the output devices were insulated - sort of defeating the whole point of the design.
Also, the Input buffer and error amp need a +/-15 Volt supply, this is generated from the 55V supplies by using a pair of big hot dropper resistors and a pair of (hot) zener diodes. This is much cheaper than adding an auxillary winding to the mains transformer but again makes the amp a pig to test.
if you run it up off a variac - for example.. there isn't enough voltage to fire up the ICs.
so you need a seperate 15V supply (through a couple of safety resitors).
Last but not least, on the AH4000, running it up through a variac meant that the fault circuit though there was a fault - so you need to bypass that too !
Repairing the unit (or at least, attempting to).
1)firstly - draw a picture of which power transistors ect are where on the board.
2) remove all the devices that are on the heatsink (makes it easier to test them and the components that were connected to them.
3) go for the output devices - check them all (out of circuit)
4) check the rail switches (probably IRF640's)
5) check the flyback didoes (T0-220 devices on the heatsink next to the IRFs)
6) check the drivers.
7)check the big resitors that were on the emmitors of the output devices. ~0.2R
8)check the driver emmitor resistors ~3R
when a semiconductor dies it usually takes its associated current limit resitor with it.
some output devices will be dead.
the rail switches might be dead.
the drivers will usually be ok.
if a driver is dead then its more likely that the fault has spread backwards through the amp.
Once you've gone through that lot, you need to check the bias ladder is ok - check for ~0.6V across 1N400x diodes check the 5v/ish zeners - if in doubt just replace them they only cost pennies.
Important ! - check the bias preset hasn't gone open circuit.. if it does, this amp will die.
Once you've replaced all the dead parts - its time to start testing the amp.
DO NOT JUST TRY IT ON 240V !! - it just wastes money!
bypass the fault circuit, run the IC's off around +/- 10 to 15V, run the mains transformer off about 60VAC instead of 240VAC (oh, you need to bypass the soft start circuit too).
Of course the rail switches don't kick in until you're above 50V on the lower rails.
not easy !
I've cant attach the circut for the QSC RMX 2450 which this amp is based on because its over the file size limit - PM me you email and i'll send it over
good luck
j
I'm afriad these amps aren't designed very well.
they are based on a circuit used by several different manufacturers including behringer and QSC.
i'm afraid this isn't a good amp to cut your teeth on.
From experience, the burned resistors aren't the cause of the fault, they are the result of several output devices failing short circuit.
your first job would be to identify the topology of the amp..
this 2400 is known as a class-H (or more correctly in this case, a Class-G - but they are quite similar).
The amp has a balanced input buffer (IC) then the sound goes to the level pot, in into the error amp (another op-amp IC).
The inputs of the error amp are:
the sound you want (from the level pot) and...
the sound you're getting (from the output of the amp, reduced by approx 26dB by the feedback resistor)
the 2 signals are subtracted from each other, and the resulting error signal is used to drive the output stage.
the error signal is also used to drive the clip led (to show too much error).
this is pretty much the same for most amps.
on this amp, the error goes into the bases of a pair of driver transistors which are in series across the high-voltage power rails, between them is the bias circuit (a few dides in series, some of which are in thermal contact with the heatsink for stability) which stops crossover distortion by making sure the opposing sets of output divices are always slightly on (with a quiescent current flowing through them).
there is a preset (~100R) there wich sets the bias current - easy to spot because it has a thermistor across it. (on the heatsink too).
There are 2 other presets ~1K - each with a diode in series with them - these limit the drive, thus limiting the output current to the speaker.
the drivers will be T0-220 devices (probably plastic 2SA1943 / 2SC4793 or metal tab MJ15023/33) these are the opposite way round to what you'd expect - so the N-Type (+ve rail) output are driven by a P-Type predriver.
depending on whether audiohead have modified the circuit from the qsc or not, there may be a second stage of driver transistors between the first pair and the output devices.
What makes this a class-g amp is that there are 2 sets of high voltage rails.
the output devices are fed from +/- 55V, but when its needed the rails are switched up to +/-110V - just for the part of the waveform that needs it.
this is done by a pair of FET transistors acting as switches whith their drive circuits sat monitoring the output stage drive voltage.
this reduces thermal wastage in the output stage and is rather clever.
the other clever feature of this range of amps is that the amp actually drives the power supply, not the speaker. difficult to explain, but it means that all the output devices have their collectors at ground potential which means they can be fixed to the heatsink without insulation.
This is great, but it means that the amp is a pig to test (the ground reference is actually the LS+ terminal and you need a floating scope (not reccomended for novices).
In the AH4000 i was looking at, AH had decided to ditch the grounded heatsink so that they could add an earth lift switch... so all the output devices were insulated - sort of defeating the whole point of the design.
Also, the Input buffer and error amp need a +/-15 Volt supply, this is generated from the 55V supplies by using a pair of big hot dropper resistors and a pair of (hot) zener diodes. This is much cheaper than adding an auxillary winding to the mains transformer but again makes the amp a pig to test.
if you run it up off a variac - for example.. there isn't enough voltage to fire up the ICs.
so you need a seperate 15V supply (through a couple of safety resitors).
Last but not least, on the AH4000, running it up through a variac meant that the fault circuit though there was a fault - so you need to bypass that too !
Repairing the unit (or at least, attempting to).
1)firstly - draw a picture of which power transistors ect are where on the board.
2) remove all the devices that are on the heatsink (makes it easier to test them and the components that were connected to them.
3) go for the output devices - check them all (out of circuit)
4) check the rail switches (probably IRF640's)
5) check the flyback didoes (T0-220 devices on the heatsink next to the IRFs)
6) check the drivers.
7)check the big resitors that were on the emmitors of the output devices. ~0.2R
8)check the driver emmitor resistors ~3R
when a semiconductor dies it usually takes its associated current limit resitor with it.
some output devices will be dead.
the rail switches might be dead.
the drivers will usually be ok.
if a driver is dead then its more likely that the fault has spread backwards through the amp.
Once you've gone through that lot, you need to check the bias ladder is ok - check for ~0.6V across 1N400x diodes check the 5v/ish zeners - if in doubt just replace them they only cost pennies.
Important ! - check the bias preset hasn't gone open circuit.. if it does, this amp will die.
Once you've replaced all the dead parts - its time to start testing the amp.
DO NOT JUST TRY IT ON 240V !! - it just wastes money!
bypass the fault circuit, run the IC's off around +/- 10 to 15V, run the mains transformer off about 60VAC instead of 240VAC (oh, you need to bypass the soft start circuit too).
Of course the rail switches don't kick in until you're above 50V on the lower rails.
not easy !
I've cant attach the circut for the QSC RMX 2450 which this amp is based on because its over the file size limit - PM me you email and i'll send it over
good luck
j
Continuing Saga.
The latest post is from another victim of these amps and he deserves our sympathy.
Fortunately, the one I repaired ( AH1000 ) didn't have the dual rail switching arrangement - it was probably too old.
All my problems were caused by an intermittant connection on the bias pot of one channel - it went open circuit as soon as there was any current flowing through it, so it measured OK with a DVM...
I've commented before that the circuit design uses this part in the wrong way - it SHOULD have been arranged so it would fail safe.
I agree with all the comments made re:- how to proceed and all I can add is that I checked my output devices with a 12 V DC supply and a car headlamp bulb to make sure they still have gain. Just checking them for open/short isn't enough. 1K from the 12V into the base should be enough to turn them on and pass 5 amps through the bulb.
As ever, good luck to anyone who has one of these units.
Graham
The latest post is from another victim of these amps and he deserves our sympathy.
Fortunately, the one I repaired ( AH1000 ) didn't have the dual rail switching arrangement - it was probably too old.
All my problems were caused by an intermittant connection on the bias pot of one channel - it went open circuit as soon as there was any current flowing through it, so it measured OK with a DVM...
I've commented before that the circuit design uses this part in the wrong way - it SHOULD have been arranged so it would fail safe.
I agree with all the comments made re:- how to proceed and all I can add is that I checked my output devices with a 12 V DC supply and a car headlamp bulb to make sure they still have gain. Just checking them for open/short isn't enough. 1K from the 12V into the base should be enough to turn them on and pass 5 amps through the bulb.
As ever, good luck to anyone who has one of these units.
Graham
Hi
All
Thanks for the replies;
AudioSystemsEng: Looks like i've got my work cut out LOL. I'm no professional when it comes to electronics but I know the basic essentials, i'll give youe list a try and i'm sure i'll be asking for more help when I get stuck, I'll try and take some pictures for you to look at.
Regards
Nelix
All
Thanks for the replies;
AudioSystemsEng: Looks like i've got my work cut out LOL. I'm no professional when it comes to electronics but I know the basic essentials, i'll give youe list a try and i'm sure i'll be asking for more help when I get stuck, I'll try and take some pictures for you to look at.
Regards
Nelix
OK, so I think I've sorted the image issue.....hopefully
Here's the internals of the amp:
And here's the offending burnt out resistors:
Any clues anyone?
Regards
Nelix
Here's the internals of the amp:
An externally hosted image should be here but it was not working when we last tested it.
And here's the offending burnt out resistors:
An externally hosted image should be here but it was not working when we last tested it.
Any clues anyone?
Regards
Nelix
OK, so now I feel a fool.........
I got this amp in this condition and upon inspecting the inside noticed the above pictured resistors and instantly thought that i had found the problem!!!!!
However....... a friend came round who is more knowledgable than me and we started a process of ilimination, during this process we discovered that it is actually the other channel that is off and there does not seem to be anything obviously wrong on the board that would indicate a problem.
Ho Hum, we are going to investigate further and I will keep you informed, i'm going to change those resistors anyway along with the same 2 on the channel thats off as they are rather black, not as bad as those above mind but better safe that sorry.
Regards
Nelix
I got this amp in this condition and upon inspecting the inside noticed the above pictured resistors and instantly thought that i had found the problem!!!!!
However....... a friend came round who is more knowledgable than me and we started a process of ilimination, during this process we discovered that it is actually the other channel that is off and there does not seem to be anything obviously wrong on the board that would indicate a problem.
Ho Hum, we are going to investigate further and I will keep you informed, i'm going to change those resistors anyway along with the same 2 on the channel thats off as they are rather black, not as bad as those above mind but better safe that sorry.
Regards
Nelix
by pure guesswork from your photos..
working from the rear of channel A you have ...
Thermal Trip
Driver (NPN) }
Driver (NPN) } in parallel 2SC4983
4 PNP output devices 2SA1943
Double Fast Diode MUR3020 or similar
a gap
then under the metal sheet..
something i cant remember,
a space for a second switching FET
the switching FET IRF640
the underside of channel A is the same as the side of channel B that is in View.
so working from the rear of channel B you have...
24V regulator for the fan circuit
The bias diodes and thermistor
a pair of PNP drivers prolly 2SA1837
4 NPN output devices 2SC3281
Fast Diode MUR again..
space (probably for more output devices of a bigger product)
The switching FET
The one i cant remember
The 2nd Thermal trip.
one thermal trip speeds the fan up - the other turns it off.
the markings on the power devices are almost invisable.
If i were doing this amp myself i'd do this...
Identify the power devices - write what the devices are on the heatsink.
also, mark the heatsinks with notes like 'channel A, top, back, front' (seriously)
draw a picture of each module showing which wires go where (there are more tabs than wires).
remove the dead module (make sure the power supply caps are discharged through a 47R / 10W resistor if you've had it powered up in the last 24hrs).
1) remove all the output devices, drivers FETs and the big diode.
2) test these out of circuit.
3) while these are off the board, it is easier to test the other components on the board.
output devices and FETs are most likely to go.
the drivers are sensitive too.
the big white emitter resistors don't usually fail in this amp.
If the drivers fail, they usually take their emitter resistors with them (something like 12R / 1W).
Other important things to check:-
The diodes in the bias ladder (2 are on the heatsink).
The bias preset.
The current limitig diodes and their presets.
(any other diodes while you're at it).
last but not least - there are at least 2 sets of voltage dropper resistors with associated 15V zeners to give +/- 15V supplies to the input and error op amps, the second gives power to the comparitors at the other end of the board that trigger the rail switches.
you need to check these power supplies are ok - the resistors are always pretty burned (crap design) - but its worth replacing the zeners 15V / 1.3W.
other burned power resistors you might find are:
the snubbers on the rail switches (a 3W resistor in series with a 100nf) to soak up the switching spike.
finally, if the amp is run at clipping with no load then there is another 3W resistor in series with something like a 100nf cap across the output. This resistor will probably be burned but still in tolerance.
you can find the values of the burned resistore by looking at the good channel (they may differ slightly from the values in the QSC diagram).
I hope this is of some help.
good luck
jonathan
working from the rear of channel A you have ...
Thermal Trip
Driver (NPN) }
Driver (NPN) } in parallel 2SC4983
4 PNP output devices 2SA1943
Double Fast Diode MUR3020 or similar
a gap
then under the metal sheet..
something i cant remember,
a space for a second switching FET
the switching FET IRF640
the underside of channel A is the same as the side of channel B that is in View.
so working from the rear of channel B you have...
24V regulator for the fan circuit
The bias diodes and thermistor
a pair of PNP drivers prolly 2SA1837
4 NPN output devices 2SC3281
Fast Diode MUR again..
space (probably for more output devices of a bigger product)
The switching FET
The one i cant remember
The 2nd Thermal trip.
one thermal trip speeds the fan up - the other turns it off.
the markings on the power devices are almost invisable.
If i were doing this amp myself i'd do this...
Identify the power devices - write what the devices are on the heatsink.
also, mark the heatsinks with notes like 'channel A, top, back, front' (seriously)
draw a picture of each module showing which wires go where (there are more tabs than wires).
remove the dead module (make sure the power supply caps are discharged through a 47R / 10W resistor if you've had it powered up in the last 24hrs).
1) remove all the output devices, drivers FETs and the big diode.
2) test these out of circuit.
3) while these are off the board, it is easier to test the other components on the board.
output devices and FETs are most likely to go.
the drivers are sensitive too.
the big white emitter resistors don't usually fail in this amp.
If the drivers fail, they usually take their emitter resistors with them (something like 12R / 1W).
Other important things to check:-
The diodes in the bias ladder (2 are on the heatsink).
The bias preset.
The current limitig diodes and their presets.
(any other diodes while you're at it).
last but not least - there are at least 2 sets of voltage dropper resistors with associated 15V zeners to give +/- 15V supplies to the input and error op amps, the second gives power to the comparitors at the other end of the board that trigger the rail switches.
you need to check these power supplies are ok - the resistors are always pretty burned (crap design) - but its worth replacing the zeners 15V / 1.3W.
other burned power resistors you might find are:
the snubbers on the rail switches (a 3W resistor in series with a 100nf) to soak up the switching spike.
finally, if the amp is run at clipping with no load then there is another 3W resistor in series with something like a 100nf cap across the output. This resistor will probably be burned but still in tolerance.
you can find the values of the burned resistore by looking at the good channel (they may differ slightly from the values in the QSC diagram).
I hope this is of some help.
good luck
jonathan
just reading your last post which came while i was writing..
on the AH4000 - I found nearly every 3W resitor was burned (mine had 2 dead channels)
the resistors i've listed above run so hot at some point they will look very burned.
i opted to replace the lot because i was doing a total re-build of the amp.
the droppers and snubbers will burn on these amps and i already mentioned about the one that burns if the amp clips.
definately replace the burned ones on the good channel.
j
on the AH4000 - I found nearly every 3W resitor was burned (mine had 2 dead channels)
the resistors i've listed above run so hot at some point they will look very burned.
i opted to replace the lot because i was doing a total re-build of the amp.
the droppers and snubbers will burn on these amps and i already mentioned about the one that burns if the amp clips.
definately replace the burned ones on the good channel.
j
I must just say that you know your stuff.
I realise, as you said, you were using guess work, heres the numbers off the devices.
Thermal Trip
Driver (NPN) }
Driver (NPN) } in parallel - only markings on these are A1837
4 PNP output devices 2SC5200
Double Fast Diode MUR3060PT
a gap
then under the metal sheet..
IRF540
gap
IRF540
Dont know if this makes a difference, I'll try check everything you mentioned and get back to you.
Thanks for all your help, it's much appreciated.
Regards
Nelix
P.S. I've removed the offending resistors and checked there values, they are colour coded as 10R with 5% tolerence. They read a little high on the multimeter at 11.3 ish is this normal?
I realise, as you said, you were using guess work, heres the numbers off the devices.
Thermal Trip
Driver (NPN) }
Driver (NPN) } in parallel - only markings on these are A1837
4 PNP output devices 2SC5200
Double Fast Diode MUR3060PT
a gap
then under the metal sheet..
IRF540
gap
IRF540
Dont know if this makes a difference, I'll try check everything you mentioned and get back to you.
Thanks for all your help, it's much appreciated.
Regards
Nelix
P.S. I've removed the offending resistors and checked there values, they are colour coded as 10R with 5% tolerence. They read a little high on the multimeter at 11.3 ish is this normal?
yep transistors often omit the "2S" from they're marking.
you're right about the 2SC5200's - i was getting my complimentary pairs mixed up.
5200 goes with 1943
on the subject of the resistors +15% wouldn't worry me.
as they burn / age thier resistance goes up, possible because the metal film spiral is getting thinner.
i'm pretty sure the pair of 10R's is attached to a 100n to make a snubber - but that's really really from memory.
just googled the IRF540 - its a lower voltage, higher current spec than the more common (and probably cheaper 640)
IRF640
200V 18A, Rds on = 0.018R
Tr = 27ns 1200pf input capacitance
IRF540
100V 33A, Rds on = 0.044R
Tr = 35ns 1900pf
so the pair of 540's can swithch ~80A, the potential across them is 55V.
use 540's if you can get them, i'd be quite happy to use 640's at a push, they are faster due to the slightly lower gate capacitance.
maybe thats why there is space for 3 FETs ?
The AH4000 used 4 x IRF640 in parallel.
you're right about the 2SC5200's - i was getting my complimentary pairs mixed up.
5200 goes with 1943
on the subject of the resistors +15% wouldn't worry me.
as they burn / age thier resistance goes up, possible because the metal film spiral is getting thinner.
i'm pretty sure the pair of 10R's is attached to a 100n to make a snubber - but that's really really from memory.
just googled the IRF540 - its a lower voltage, higher current spec than the more common (and probably cheaper 640)
IRF640
200V 18A, Rds on = 0.018R
Tr = 27ns 1200pf input capacitance
IRF540
100V 33A, Rds on = 0.044R
Tr = 35ns 1900pf
so the pair of 540's can swithch ~80A, the potential across them is 55V.
use 540's if you can get them, i'd be quite happy to use 640's at a push, they are faster due to the slightly lower gate capacitance.
maybe thats why there is space for 3 FETs ?
The AH4000 used 4 x IRF640 in parallel.
Has anyone managed to get an AH4000 to work and STAY WORKING !
I have one here that i did a lot of work on. but it had a tendancy to blow up for no apparent reason on the test bench.
14 output devices per channel means i'm not going to throw any more money at it.
I'm lacking information on setting the bias (i used a rule of thumb of 18mV measured across an emitter resistor).
I aslo need a reliable way to set the current limiting, and any information on the correct shape of the rail switching waveforms.
one rail is a neat square wave... but the other had a long 'off' tail. - is this normal ?
My gut reaction is to steer clear of the audioheads - if they were a truley reliable amp then there wouldn't be so many different modifications.. oh, and audiohead would still be in business ;-)
j
I have one here that i did a lot of work on. but it had a tendancy to blow up for no apparent reason on the test bench.
14 output devices per channel means i'm not going to throw any more money at it.
I'm lacking information on setting the bias (i used a rule of thumb of 18mV measured across an emitter resistor).
I aslo need a reliable way to set the current limiting, and any information on the correct shape of the rail switching waveforms.
one rail is a neat square wave... but the other had a long 'off' tail. - is this normal ?
My gut reaction is to steer clear of the audioheads - if they were a truley reliable amp then there wouldn't be so many different modifications.. oh, and audiohead would still be in business ;-)
j
Funnily enough I had one last week where the customer had brought it in saying the left channel was faulty and he'd been using it ok just on the right channel. I fixed the left channel (blow up) and put in on test where I found the right channel was badly distorted. I replaced the rail switch snubber resistors (open circuit) only to see them go open again; I scraped the varnich off the snubber caps and discovered they were leaky..I upgraded them to some nice Philips polyester ones that did the trick. But I still had the distortion..the scope showed it to be crossover distortion, a simple tweak of the bias pot cured it.
Without knowing the correct bias measurement, I usually turn the bias pot (where there's only the one) to minimum bias (ie max crossover distortion) then advance it slowly till the distortion just disappears on a sinewave of 2kHz into a dummy 4 ohm load running at low output. It should of course happen around the pot mid-way mark but dont quote me on that.
I've never had to replace any current limit bits on the AH units so can't help there but it usually involves running the amp into a short circuit load and setting for amplitude reduction at a particular current consumption..sometimes you have to short out the bus fuses as well during the setting.....not for the weak hearted.
Without knowing the correct bias measurement, I usually turn the bias pot (where there's only the one) to minimum bias (ie max crossover distortion) then advance it slowly till the distortion just disappears on a sinewave of 2kHz into a dummy 4 ohm load running at low output. It should of course happen around the pot mid-way mark but dont quote me on that.
I've never had to replace any current limit bits on the AH units so can't help there but it usually involves running the amp into a short circuit load and setting for amplitude reduction at a particular current consumption..sometimes you have to short out the bus fuses as well during the setting.....not for the weak hearted.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.