I have a B&k 4420 amplifier that has low output on one channel. Here's what I checked so far. Dc offset at speaker terminals it is good and adjusted to 0 mv. It does jump around a little maybe 5 mv or so. I checked the power supply voltage all good. I checked bias with 100mv 20khz sine wave input and set it with a meter in place of the fuse and it's set to 250ma. One channel with the 100mv rms input puts out 2.15 volts into 8 ohm dummy load. The bad channel only puts out 1.09 volts. I have a scope and a signal generator, dummy loads and a variac. I did check all emitter resistor voltage and there are a couple that are way higher than the rest. I am at work right now and don't have the exact numbers but I can post them later. I do have a schematic and can post it if there is no problem with it. I hope someone can point me in the right direction, I'm in the learning process a bit and could use some help with this one. Thanks in advance. Mike
After the basics, above, one usually checks AC voltage through the circuit to see where the imbalance starts. Scope is good. Signal gen good but Y the output to both channels so signal is same to both.
New amp, solder joints are suspicious. 5 year or older amp,electrolytic or tantalum coupler caps, volume pots, punch down connectors like ribbon cables, etc, connections like op amp sockets. Oxide affects the latter three, remove and replace or spray. Look at input, after blocking cap, then collector or output of 1st stage and VAS. Drivers you can't see much in the voltage realm, look across resistors with floating ground drivers & output. Voltage across resistors determines current, which should be balanced in both channels. current should be balanced also across all emitter resistors of multiple output transistors.
Output stage problems usually follow a wiring problem in amps used on stage or moved around. I don't spend much time on output stage at first, but commercial repairmen serving the bar band market always start there. Stumbling around on stage in the dark with too many 1/4 phone plugs leads to a lot of problems for small venue bands.
New amp, solder joints are suspicious. 5 year or older amp,electrolytic or tantalum coupler caps, volume pots, punch down connectors like ribbon cables, etc, connections like op amp sockets. Oxide affects the latter three, remove and replace or spray. Look at input, after blocking cap, then collector or output of 1st stage and VAS. Drivers you can't see much in the voltage realm, look across resistors with floating ground drivers & output. Voltage across resistors determines current, which should be balanced in both channels. current should be balanced also across all emitter resistors of multiple output transistors.
Output stage problems usually follow a wiring problem in amps used on stage or moved around. I don't spend much time on output stage at first, but commercial repairmen serving the bar band market always start there. Stumbling around on stage in the dark with too many 1/4 phone plugs leads to a lot of problems for small venue bands.
Does your amp version have input level controls, or instead balanced inputs?
If it has balanced inputs, check the push button next to the inputs, because
it may be dirty or in the wrong position. I think it's the switch, because the
difference between the channels is a factor of two.
Last edited:
Ok I haven't had both sides running together yet but I can do that and see where the voltages change. I do know that on each side there is one emitter resistor that measures quite a bit higher than the rest of them in line. It is the model with the xlr and it has a switch that I was checking out last time I was working on it , I will look into that more. I don't think it's the pots because they seemed rather accurate even though they are one turn pots. I will clean any connectors and do some more measurements and post back what I find. Thanks for the help, I appreciate it very much.
I'd replace the emitter resistor that's off in value, regardless. Clean and exercise that
balanced mode on/off switch, it should help.
For multiple output devices, could be. It's also common for the emitter resistors to be
off in value or even open. Check their resistances with a DVM (with power off).
Ok I did play with the switch lubed it and worked it quite a bit and still the same. I checked the emitter resistors with my meter but I couldn't get a definite reading as they are .47 ohm too low for my meter but they did all read about the same. Tomorrow night I will mess around with the scope and try to check across various resistors and compare the two channels.
Ok I measured the D.C. Voltage across every resistor on each board when the amp was warm. They were all pretty much the same give or take a little bit. This was with my multimeter. Now I should check the ac or signal throughout the circuit with my scope? How exactly should I do this ? I would assume I ground the clip and measure each resistor with the probe? Or do I check at the transistors? Thanks for any guidance. I don't think it is the switch because the signal coming out of the switch board is the same as what's going into it.
The circuit DC voltages using your DVM will likely show the problem.
Compare the two channels and see if any DC voltage is substantially
different in the bad channel.
Never connect a scope ground clip anywhere but ground, regardless.
What I do on a sound imbalence problem, is spot the input op amp or transistor, the VAS transistor,the driver transistors. Then I check AC voltage at the collector resistor of 1,2,3,. You may have an input problem that would show up at the input of first op amp or base of first transistor. Voltage checks you can do with scope probe ground on analog ground, except Peavey & other bridgeable amps with flying speaker ground.
If there is an oxidation or input e-cap problem, there won't be any DC voltages off.
You spot the transistors by size. little plastic ones near the input, bigger ones in the middle like TO-126, TO220's with heat sink tabs and probably heat sinks for drivers, going out to the big TO3 or TO3P output transistors. Collector resistor would be the intermediate value usually. High resistance on base, low resistance on emitter. A picture printed out would help to write voltages down on, if you have a cell phone & a printer. then if you measure the wrong thing, time is not wasted.
If the AC is halved before some point between channels, the problem is before that on the low value channel.
If you have the schematic diagram with parts layout, even better of course. Put your amp name & schematic in a search engine, then ignore tha pay sites that come up first.
Proper scopes like Tektronics & HP have a floating probe ground, isolated from the chassis ground safety ground or power supply ground. Some scopes sold to newbies and amateurs do not. I'd do any current checks around the drivers & output transistors last, on an AC imbalence between channels problem. Current checks with an ungrounded probe scope, you check voltage across a resistor. I'd check probe ground isolation with a DVM versus case & safety ground, even on a Tek or HP scope I bought used.
"PC scopes" sold for the sound card can ground to the computer digital ground. This can blow up your PC. I wouldn't use one of those on any amp project except possibly at the speaker output once it was proved working correctly.
As far as DC checks, you look for the impossible. Voltage across a resistor higher than the wattage of the resistor would allow. No DC voltage across a capacitor. Transistor Vbe not 0.55 to 0.7 v. collector voltage not between 1/3 and rail voltage. Op amp pins not all same DC value, in +, in -, out. Diodes not 0.55 to .7 v forwards or some DC voltage backwards.
BTW I don't own any instrument that can tell the difference between a .22 ohm resistor and a .47 ohm resistor. If I use a known calibration resistor, sometimes my Sears 200 ohm scale DVM will show a difference between a 2 ohm resistor and a 1 ohm resistor. That's only with a new battery, a tired battery won't even show that. More accuracy than that, buy a special low ohms meter for $$$, or build a wheatstone bridge.
Best of luck.
If there is an oxidation or input e-cap problem, there won't be any DC voltages off.
You spot the transistors by size. little plastic ones near the input, bigger ones in the middle like TO-126, TO220's with heat sink tabs and probably heat sinks for drivers, going out to the big TO3 or TO3P output transistors. Collector resistor would be the intermediate value usually. High resistance on base, low resistance on emitter. A picture printed out would help to write voltages down on, if you have a cell phone & a printer. then if you measure the wrong thing, time is not wasted.
If the AC is halved before some point between channels, the problem is before that on the low value channel.
If you have the schematic diagram with parts layout, even better of course. Put your amp name & schematic in a search engine, then ignore tha pay sites that come up first.
Proper scopes like Tektronics & HP have a floating probe ground, isolated from the chassis ground safety ground or power supply ground. Some scopes sold to newbies and amateurs do not. I'd do any current checks around the drivers & output transistors last, on an AC imbalence between channels problem. Current checks with an ungrounded probe scope, you check voltage across a resistor. I'd check probe ground isolation with a DVM versus case & safety ground, even on a Tek or HP scope I bought used.
"PC scopes" sold for the sound card can ground to the computer digital ground. This can blow up your PC. I wouldn't use one of those on any amp project except possibly at the speaker output once it was proved working correctly.
As far as DC checks, you look for the impossible. Voltage across a resistor higher than the wattage of the resistor would allow. No DC voltage across a capacitor. Transistor Vbe not 0.55 to 0.7 v. collector voltage not between 1/3 and rail voltage. Op amp pins not all same DC value, in +, in -, out. Diodes not 0.55 to .7 v forwards or some DC voltage backwards.
BTW I don't own any instrument that can tell the difference between a .22 ohm resistor and a .47 ohm resistor. If I use a known calibration resistor, sometimes my Sears 200 ohm scale DVM will show a difference between a 2 ohm resistor and a 1 ohm resistor. That's only with a new battery, a tired battery won't even show that. More accuracy than that, buy a special low ohms meter for $$$, or build a wheatstone bridge.
Best of luck.
Last edited:
Ok so while probing with my scope around the first small signal stage I he two different types of waves from the good and bad side. Now I have never hooked it up to speakers or even really gave anything higher than 100mv sine wave, but I assume the louder channel is the good one. The bad channel (lower output) or maybe their both bad not sure looks like a saw tooth when zoomed in looking at maybe an oscillation. And it's at a lower amplitude than the louder channel. The louder channel looks like an X or double saw tooth. And it's at a larger amplitude. There are for sure some differences between the signal in the first stage.ill try to post the pics of the scope traces. I can't seem to post the pics from my phone, any ideas?
Well, great. If you have channel imbalence, then one looks at the softer channel, the one with lower voltage.
If the problem is at the base of the input transistor, you have a problem with oxidized connectors, oxidized or dropping out input volume pot, or a bad coupler cap. The base is the middle leg of most transistors, the left leg looking at the flat on others. There will be 0.6 to 0.7 vdc b to e. The EBC ones the e is the left., the BCE ones the E is the right. Use a DVM when poking around there usually, the big scope ground clip is too big to use poking around at leads.
If a oxidized connector power off remove & replace. If oxidized pot wiper, wiggle back & forth or spray out with contact cleaner, or replace. If bad input cap (usually 1 uf to 10 uf) replace.
If the music (or 100 mv signal source) goes in the base okay but comes out the C wrong, then look at DC voltages. B to E 0.6 or .7, C to E should be 1/4 to 3/4 of the input stage Vcc. some amps drop the Vcc input with a resistor from the rail voltage, some use constant current sources (very complicated arrangements with other transistors).
Telling which little transistor is the input amplifier, and which is a constant current source, or the feedback amp of a differential pair, takes a little circuit tracing, or looking at the schematic and parts layout diagram, ideally. Sawteeth could be the constant current circuit at work. You should see sine waves around the input cap and before the resistor to the input amp transistor base.
You picture didn't work, but I don't really need it. I don't use a scope at home so I'm not an expert on amp waveforms using them. At work, I didn't work on amps, I worked on A/D data acquisition production test equipment.
Note on old equipment, the freight from the parts warehouse, or the gas to drive to Fry's store, is 20 x the cost of the cap, usually. If one is bad, I buy all the electrolytic caps, not just the "one that is bad". Then I replace the one that is bad, get the amp working, then replace the others 2 or 3 at a time with a test in between, so I can make sure I didn't make a bad solder joint or bad parts substitution which is causing a new problem. Rubber in e-caps deteriorates running or sitting on the shelf, except for the very rare, almost mythical, epoxy sealed e-cap. I've had to replace e-caps in one amp 4 times in 46 years, so now that we have debit cards & the internet, I buy "long life" e-caps, the ones with 3000 to 10000 hours service life, the higher I can get the better. All distributors show you the **** grade caps first, they are cheaper, and most repair services won't use anything better.
If you replace one e-cap, the one that is bad, your amp is broken all the time as the others fail one by one. Gives used equipment a bad name. Around here, to look at used equipment I've bought, the pro repairment put used ecaps in equipment. That makes sure the equipment will go bad again, and soon. I found a 1981 cap in a 1998 amp, last week. This kind of pro work makes amateurs stumbling around learning things for the first time look good. Used resistors if checked, and film or ceramic caps, are okay, especially if you read the one you're putting in with a meter. They don't age much.
If the problem is at the base of the input transistor, you have a problem with oxidized connectors, oxidized or dropping out input volume pot, or a bad coupler cap. The base is the middle leg of most transistors, the left leg looking at the flat on others. There will be 0.6 to 0.7 vdc b to e. The EBC ones the e is the left., the BCE ones the E is the right. Use a DVM when poking around there usually, the big scope ground clip is too big to use poking around at leads.
If a oxidized connector power off remove & replace. If oxidized pot wiper, wiggle back & forth or spray out with contact cleaner, or replace. If bad input cap (usually 1 uf to 10 uf) replace.
If the music (or 100 mv signal source) goes in the base okay but comes out the C wrong, then look at DC voltages. B to E 0.6 or .7, C to E should be 1/4 to 3/4 of the input stage Vcc. some amps drop the Vcc input with a resistor from the rail voltage, some use constant current sources (very complicated arrangements with other transistors).
Telling which little transistor is the input amplifier, and which is a constant current source, or the feedback amp of a differential pair, takes a little circuit tracing, or looking at the schematic and parts layout diagram, ideally. Sawteeth could be the constant current circuit at work. You should see sine waves around the input cap and before the resistor to the input amp transistor base.
You picture didn't work, but I don't really need it. I don't use a scope at home so I'm not an expert on amp waveforms using them. At work, I didn't work on amps, I worked on A/D data acquisition production test equipment.
Note on old equipment, the freight from the parts warehouse, or the gas to drive to Fry's store, is 20 x the cost of the cap, usually. If one is bad, I buy all the electrolytic caps, not just the "one that is bad". Then I replace the one that is bad, get the amp working, then replace the others 2 or 3 at a time with a test in between, so I can make sure I didn't make a bad solder joint or bad parts substitution which is causing a new problem. Rubber in e-caps deteriorates running or sitting on the shelf, except for the very rare, almost mythical, epoxy sealed e-cap. I've had to replace e-caps in one amp 4 times in 46 years, so now that we have debit cards & the internet, I buy "long life" e-caps, the ones with 3000 to 10000 hours service life, the higher I can get the better. All distributors show you the **** grade caps first, they are cheaper, and most repair services won't use anything better.
If you replace one e-cap, the one that is bad, your amp is broken all the time as the others fail one by one. Gives used equipment a bad name. Around here, to look at used equipment I've bought, the pro repairment put used ecaps in equipment. That makes sure the equipment will go bad again, and soon. I found a 1981 cap in a 1998 amp, last week. This kind of pro work makes amateurs stumbling around learning things for the first time look good. Used resistors if checked, and film or ceramic caps, are okay, especially if you read the one you're putting in with a meter. They don't age much.
Last edited:
Ok I think the problem is the electrolytic caps. The two big 100uf 100volt ones are shot. I took them out and checked them on my scope and have super high esr. So I have all new Panasonic low esr caps to put in and I also have new multi turn pots to replace the original ones. One thing I noticed is that the 2n5210 transistors are made by different companies throughout the amp. They are the same for each channel though. Q3,4 are marked with a fancy N 606followed by 2n5210. The Q1,2,5,6 are marked CEN 2n5210. Then Q7,8 are CEN 2n5087. They look to be original. I will replace the electors and sealed pots and see if that does the trick. I'm pretty sure it will since those caps were basically doing nothing. After I get it working properly would this amp benefit from replacing the lot transistors with some matched pairs and tie them together with some heat shrink over each pair? One more question I have a bryston 4b st on my bench that had a broken pcb trace the lip transistors are marked on the board as being the same as the B&k except all or most transistors are marked with an M probably Motorola. But the ltp trans are totally different they are F pn200a and pn100a. I'm thinking Farnel transistors? Why would B&k use different brands of transistors throughout? And are the pn100a and pn200a equivalents or better? What transistors could I find easily to match closely to replace the 2n5210? Thanks a lot for all the help and tips.
I wouldn't worry about B&O purchasing practices. Bang and Olovsen of Sweden, right? The people that used to sell the $400 calibrated flat microphones to science? They would have an incoming inspection department, and probably match long tail pairs and match current mirrors with their own equipment, not worrying about who made the part.
If a transistor has .6 to .7 v on the BE, and higher voltage on the collector (or minus for pnp) I wouldn't worry about it. Bad solder joints and oxidized connectors are more of an issue than transistor failure except at the output where wiring and speaker faults can blow transistors up.
I would post the schematic and let the experts look at it before "improving" anything. Even mostly sure things like paralleling >100 uf electrolytic caps with a 1 uf film cap, can mess things up on some amps. The only thing I'm trying to improve on my PV-1.3k is the DC protection circuit. The current protection circuit melts the traces on the board to the triac and blows the triac, instead of blowing the breaker like it should.
I'd build another amp before trying to "Improve" something as prestigeous as a B&O. Start looking for salvage heat sinks and enclosures, or dead ****y amps with same. Take a look at projection TV's and LED TV's on the curb, salvage the power boards and leave the projection and display stuff for the garbage. I've found nice heat sinks diode and power resistors in them. Also a 2SC**** VAS candidate in the color circuit (150 mhz Ft, Vceo>200v) but it was blown. Can't buy good VAS here with 2SC 2SA numbers, just NTE "replacements" with no Ft spec or Cob spec. I'm waiting now for some KSC# KSA# fast VAS transistors that newark/farnell imported for me from somewhere. Going to cost me $10 freight for $8 in parts, if they don't hit me with a $20 foreign warehouse charge.
BTW as I might have said, once the amp sort of works right, every electrolytic cap and tantalum cap needs to be replaced, 2 at a time to focus your eyes on the bad solder joint or assembly error you just made. The rubber seal is as bad in all the others as the first ones to dry up. And tantalums are a curse, I try to get rid of them if I have any, 11 out of 12 I bought in old equipment have been bad.
If a transistor has .6 to .7 v on the BE, and higher voltage on the collector (or minus for pnp) I wouldn't worry about it. Bad solder joints and oxidized connectors are more of an issue than transistor failure except at the output where wiring and speaker faults can blow transistors up.
I would post the schematic and let the experts look at it before "improving" anything. Even mostly sure things like paralleling >100 uf electrolytic caps with a 1 uf film cap, can mess things up on some amps. The only thing I'm trying to improve on my PV-1.3k is the DC protection circuit. The current protection circuit melts the traces on the board to the triac and blows the triac, instead of blowing the breaker like it should.
I'd build another amp before trying to "Improve" something as prestigeous as a B&O. Start looking for salvage heat sinks and enclosures, or dead ****y amps with same. Take a look at projection TV's and LED TV's on the curb, salvage the power boards and leave the projection and display stuff for the garbage. I've found nice heat sinks diode and power resistors in them. Also a 2SC**** VAS candidate in the color circuit (150 mhz Ft, Vceo>200v) but it was blown. Can't buy good VAS here with 2SC 2SA numbers, just NTE "replacements" with no Ft spec or Cob spec. I'm waiting now for some KSC# KSA# fast VAS transistors that newark/farnell imported for me from somewhere. Going to cost me $10 freight for $8 in parts, if they don't hit me with a $20 foreign warehouse charge.
BTW as I might have said, once the amp sort of works right, every electrolytic cap and tantalum cap needs to be replaced, 2 at a time to focus your eyes on the bad solder joint or assembly error you just made. The rubber seal is as bad in all the others as the first ones to dry up. And tantalums are a curse, I try to get rid of them if I have any, 11 out of 12 I bought in old equipment have been bad.
Last edited:
It is actually B&K Components EX4420 and I made an attempt to post the S. manual by way of an attachment. I think it went so its available to view. I agree with not adding components unless you know how the circuit works and I absolutely do not. But I would take advise from others that would know. If you look at the material list they have an M next to some transistors which probably Means matched and there is a U next to some that probably means unmatched. There are a few blank capacitor spots on the boards so maybe they left out capacitors to save money(I doubt it though). Im open to opinions for any updates that could be done. Check it out and let me know. I am about to clean up the board and install the new caps and pots, Ill let you know if that fixed it or not. Thank you much.Mike
Yes it was used and one channel is half the output. I got it and a broken B&k cs115 that needed a transformer for 80$. I haven't heard it on speakers but I can't wait. I also found a B&k Ex4420M for next to nothing, waiting for it to arrive yet. It should be the same thing just with each channel bridged together. It blows the fuse on power up but we'll see. Worst case it's got some good parts and a chassis for a diy amp. I'd love to do a diy project it would probably help me understand how things work a bit better.
Ok, some good news. The bad channel is now good with the new caps. Every cap was shot and the channel I thought to be good was also bad. With a 100mv rms sine wave input I'm getting 2.8volts out. The channel I thought was good only had 2.1 or so volts output. I should have just put new caps in it from the get go but it is all part of a learning curve. Oh well off to do the other channel and then crank it up! Can't wait to see how this thing sounds compared to my adcom 555 that I think sounds pretty darn good. I don't doubt there's better but for these pretty cheap components you really can't loose. Especially when you pick them up for next to nothing broken and can get them working again. ��
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.