This was a fun one to tackle. Direct coupled amps are great until something goes wrong, and then they are a pain in the ass to diagnose, especially when something goes intermittent.
Those top half output devices are are NOT P-MOS they are N-MOS by symbol. Arrows are used in opposite direction than in BJT emitters when for FET gates. (27 minutes video point).
I've worked on a few of those amps and the output VFets are pretty much bullet proof, the signal transistors can be a problem. For it's time it was a good reliable amp that had a good sound, better than a lot of it's generation.
No matter how many times I see your videos I’m always amazed at your talents and skill. I know absolutely nothing about electronics but your videos are very entertaining. Thank you so much for posting.
I agree. He's a master at his craft and his humorous adds in post and his cat crack me up. I suspect he pays the mortgage but the cat owns his house. Kinda like here. I had to put up a child gate to keep my cats out of the workshop haha
Dave, I think you're wrong about 3N3906 being suitable for Q4. E and B are up near almost full +DCV, but the collector is held at ~+0.6 V by the lower input pair during normal operation (not counting fault conditions). This is a 100wpc amplifier, so must be able to output +/-40 V peak. One guy measured about +/-59 V supplies, which is in line with what I'd expect. Clearly *not* a place for a transistor rated only 40V C-E like the 2N3906 (the 60 V is Vebo, not Vceo!). Use a BC556B turned around if you must (rated 65 V), but another 2N5401 would really be the best bet. 2N5401 and 2N5551 are fairly common small-signal types in amplifiers this class, can't hurt to have a small stash of 'em. (Below that, BC556/546 are not uncommon.) I would add some KSC1845 / KSA992, as high-beta high-voltage transistors are getting harder and harder to obtain. 2SC2240/2SA970 have been EOL for a good while now. TO-92 is going the way of the dodo, and things are not looking up for transistors for analog applications. There is not one ultralow-noise highish-beta BJT (think 2SB737) left in production. If it weren't for Linear Systems, the situation for low-noise JFETs would be quite dire, too (RIP BF862). This reminds me of the CLASSÉ amplifiers that apparently shipped with the two transistors in a cascode reversed right from the factory - apparently a drafting error that got transferred onto the board. Many made it for decades with a small-signal part operated way out of spec. Apparently, MPSA56s are *tough* . If you ever get a DR-25/25 or similar in, check the Q6-7-9-10 area for schematic accuracy and sanity.
Hi, that control you keep referring to as a "balance" control looks like it is there to set the standing current in the output transistors. So more like a "bias" control. It is very similar in operation the the one that you normally see controlling the temperature sensing transistor (Q9) on normal bipolar output stages, just a little further back in the chain. As mosfets don't generally suffer from thermal runaway like bipolars I suppose the setting is not too critical as log as you have enough standing current to stop crossover distortion. As the amp is DC coupled and the input is a differential arrangement the DC offset should be self setting... Q4 & Q3 are setup as constant current sources for the differential pairs so an offset here will really upset the standing conditions.. Keep producing the vids. :)
Very Good work. Straightforward explanations throughout. Years of experience shine through in your presentations. No background music , no nonsense. Thankyou for making the effort to share these projects.
This is one of the reasons I sold my Adcom GFA-555. No speaker protection relays - not worth destroying your expensive speakers if something pukes in the amp.
I have a similar problem on a 1976 Kenwood KA-3500. Replaced just about everything on the power board. Including bias pot. Replaced. Still have DC. Very tricky to troubleshoot the vintage transistor amps. Xmas tree topology . 39:27
A friend of mine has one of These Hafler DH200 and he said that the left channel the sound started going down until he did not have any sound from that channel what would cause this problem? anyone know? Thank you
Hafler should have had a DC sensing relay circuit like most modern amplifiers, to disconnect the speakers if DC is present at the output. His earlier designs like the Dynaco Stereo 120 had capacitor coupled outputs which isolated the speakers from DC.
Great video as always. I've been looking for an esr meter like yours and I can't find one. All the new ones think they need bells and whistles but I just want the one like yours for a big recap job for a piece of personal audio equipment. You should add patreon. I'd join for 5 a month.
The Blue ESR meter is the one I have. I've had it for about six years. It works great, and it's not complicated to use and it's very fast. It cost $100 in kit form, and was easy to build. The only thing I really don't like is it has a digital readout, and I prefer analog. The other option is to search for ESR meter build on RU-vid and there are several that work just fine. It is a bit more work though and a bit of mechanical engineering and parts searching to get everything put together. You will also need some perfboard, etc.
@@waltschannel7465 awesome, I have all that stuff. Searching Amazon yields overpriced crap. There's a uni-t one that a bunch of sellers rebrand for ridiculous prices all over the spectrum and it still says uni-t on them lol I found a simple one from China for 55cad that I might buy. An esr tester and a decent awg are pretty much the only pieces of kit I don't have. I have a line on a decent used awg I think. I can fix video game consoles and other simple electronics but audio is new to me. Thanks alot man, being kind of a newbie I hate asking questions that people are probably sick of answering. I'm a software guy mostly because of my schooling but I regret not getting into electronic engineering/repair instead.
Great video, your troubleshooting techniques and experience shows up. Thanks for this. If you have some time, I would like to know your opinion about this: I got a Hafler HD200 with a fuse burned. After changing it i connected it to a pair of 8 ohms test speakers, it sounds great. However one side is running cold and the other gets warmer. I checked the DC output and the cold side is 195mv and the warm one is -65mv. My current assumption is that the issue is on the warm side. I have checked the output transistors and their resistors according to your process and they are ok. I also checked the input transistors and seems to be ok. Any ideas where to check? I checked all the resistors and the diodes and they are ok. It hasn't been recapped and I haven't checked the 3 transistors with the heatsink on top. What do you suggest to do next? Thanks in advance
Thank you 12voltvids for the really good videos. You show us your experience and knowledge and your delivery is just right. I am a hobbyist working on my own Hafler DH-220. It had -36V DC on one channel. This video was very helpful. Took a week of probing, head scratching, study of the schematic and more than a few youtube videos to help me find the problem. I fixed it by replacing Q1, R8 and R9. along with Q2, these provide the bias current for the input diff amps. I set up the bias(AOK at 275 mA), and the offset (less than 5mV), and load tested it on an 8 Ohm resistor bank for 30 mins or so. The amp produced 165W/channel just below clipping. It sounded OK on my benchtop speakers I was ready to celebrate my success and put if back in my system, but........ My final look at the outputs showed some fuzz. ARRGH. I have about 20mV of 96 MHz on the outputs. It is everywhere: both channel outputs and power supply. This amp has been re-capped and I am wondering where to start. Any suggestions?
It was crazy they had such elaborate transistor testers back in the 50s, 60s, and 70s. I still have my RadioShack tester from the 70s and that one was just a simple one. Now we just check junctions and for shorts.
For MOSFETs, the arrow represents the substrate->channel diode, just like on a regular PN junction. N-FET positive substrate -> negative channel. The arrow points towards the channel. Btw. I wouldn't risk my speakers without a relay protection circuit.
I just rebuilt one of these a few weeks ago, someone got all happy with the bias pots and decided to crank them all up. Fixed the issue & replaced a few other parts. DH200 good amps. Jfets are very nice !!
Why? It's not like it hasn't been powered up recently. Dim bulb is more for tube gear to protect transformers and tubes during power up. I knew it wasn't shorted. If it had cone in with a blown fuse I would have, just to save a fuse from blowing.
I have a Hafler DH-200 that I bought new (assembled) in the 1980s. A few weeks ago I discovered that the right channel doesn't output anything, Is this something you would be willing to fix for a fee?
hello 12volts, I have an old Peavey Bass amp that's in great need of repair. It is a Century 120 bass series amplifier, built in July of 1976. The problem I'mhaving is low level distortion with a very small loadI have had much distress in getting a schematic for both the preamp, and the power amp. Peavey has been very little help on this issue! I don't think they have anything thst old. Not good business practice in my opinion. Anyway, I would like to lnow if you could take some time researching this for me, in order to see how hard to fix this thing, or if it is worth fixing. I really would like to have it repaired, as I had a head similar to it some 30 years ago.
14:00 all bad solderings imo, or been heated, this is how these chinese garbage are assembled these days. between the green and white cable, a part of the pcb print has vanished? So this might already had "service" before?
I built a DH 220 about 1982. The kits came with the main left and right circuit boards assembled, tested, and biased. The kit builder was responsible for installing the wires to the circuit board that you mentioned have poor solder joints. Not Hafler's fault. The kit builder also had to assemble the power supply components and all the input and output connections.
Liked the video but someth9ng really struck me in this vid. The tight shot with amp and your hands fiddling along. It hit me how amazing the human hand is and want magnificent tool it is and all the motions it can perform. I will have to remember to use ESR meter to measure small value resistors.
Those transistors are not cheap. Hermetically sealed aren't they? Whats the rule, mosfets have infinite gain unless turned off because it can lead to thermal breakdown and they become permanently open
the dh-200 is a kracken a monster of a amp the circutry seems so simple i read that its output transistors are Hitachi mosfets? wow be ware when things fail in it the monster will cook your speakers lol i learn more about stereo equipment every time i watch your videos .
Fascinating watching a skilled technician who actually knows how to troubleshoot and use deductive reasoning. I wish we had someone like you in my area.
No current. Under load it was very low. Essentially just the bias was gone that keeps the negative MOSFET in conduction. Result very distorted sound but no real current. The 27 volts came up under no load. Now if an output shorted and dumped the full dc across the speaker that would be a worst case senerio but that is what the 5 amp speaker fuse is for so the speakers should be ok. This one with the bigger fuses though, I don't know if the power supy can sustain 7.5 amps so the fuse may not blow. The owner has been advised and the fuses will be replaced with correct value. Actually the rail fuses would probably pop if the output shorted.
It's well within the operating voltage of the circuit. I can only put in what I have access to and many of these originals are no longer available. Or if someone has one they want the moon for it. The vast majority can be subbed with absolutely no issues whatsoever.
Go look at the schematic, with all the resistors in series with the rail voltages. Max E-B voltage is 2.1V. Look at the way the 3 diodes are arranged. That sets the max voltage drop at 2.1V I have been subbing transistors for 40 years and have never run into an issue. The designer chose that value not because of the voltage, requirements but likely because it was readily available and had the hfe required. There are situations where the voltage rating is important, but in this circuit this is not one of them.
No problem. Look at the location. The way it is wired it never sees more than a couple of volts. Suspect the reason that original was used was cost. They used the same transistor for all the small pnp and did the same with npn. Cheaper to just use the same transistor throughout rather than have many different specs. If it was sitting across rail then yes you want the HV rating. This is not the case here. There are resistors between the b+ and diodes to set the max drop across the e-book junction at 2.1v. it is the voltage across the junction that kills the transistor and in this case that is clamped by the 3 in series diodes across the eb junction.
I tought that the big capacitors (at the end of the amplifier) role is to block the DC voltage and to pass the AC (music signal) to the loudspeakers.So how can I measure a DC voltage on the loudspeaker terminal with those capacitors in place and good shape?
Loot at the way the current path is articulated by the "bus bar" between the two soup can capacitors, this is because they are part of the power supply or if you will, filter caps.
@@catsbyondrepair Would have to be a lot to get me to do it. About 35 years ago I built a few 5 watt FM transmitters for a small community so they could rebroadcast radio received from satellite into the community as they had no local radio. When I was approached I said no. more than once. When they said there is a grand in it for each one, then I said OK. Bought a couple of FM1 kits from Ramsey, and built an amplifier to go along with it.