trouble with Denon POA 2800 amp - Anatech RSVP

Hi bekim,
Thank you.

Did you repair your amp yourself? It sounds like you did. Good show!
What you have noticed is very common with all consumer product, as mentioned. I was curious as to what method you used in your work.

The way I do solder connections is very time consuming. I don't expect a working technician to follow the same procedure simply due to time constraints. I add a touch of fresh solder to the connection, then remove all solder with a suction pump (not the small ones). If the lead is oxidized as it usually is, the lead is cleaned by scraping the surface with my bent pliers that have some cross-hatching near the tip. Not too hard, but enough to get to the metal. Some liquid solder flux is applied sparingly to the lead, then the solder pad. Then the component or wire is resoldered and the flux residue cleaned up. There are times when the component has been heated too much, forcing a replacement. The characteristic failure due to heat is a reduction in resistance (or capacitance), followed by a steady increase in resistance (or continued drop in capacitance with possible electrolyte leakage). It is possible for a failing resistor to measure very close to it's proper value even though it's well on it's way to going very high in resistance. Overheated transistors typically have increased leakage currents and reduced gain. Therefore, a close observation of the condition of these parts may be more important than measuring them.

I only mention this because I've seen many repairs return, failed, from otherwise excellent technicians. It's something that an effort to save a customer some money by not replacing parts will backfire and cause more trouble for everyone. Parts are far less expensive than labour is, or your time if you are repairing your own unit.

These comments are more for other readers since I have the feeling you are already aware of this.

It's always great to hear of another person who takes care in working on things.

-Chris 🙂
 
Hi anatech,


I purchase POA-2800 without many parts on it; mosfet drivers, output drivers, plenty resistors and some small electrolytes 'puffed-up'.
The most challenging was diodes, zener and standard ones, because I couldn't find any info what specs are. It tooks me about three months to put it all together, and since then it works like charm. Just, I never could find the posistors / thermal resistors...
I have Onkyo M-508 and output transistors and drivers are similar like in POA-2800 but sound wise, is different.
I enjoy listening sound from both of them.

As you said in previous post that, people who repair audio equipment especially amps, they don't know much in repairing vcr's and tv's and vice versa.... I know many people who repair TV's ;-)

Repairing techniques and tools are pretty complexed when it comes to audio equipment.
Some people repairing audio system take as granted, very easy, not respected work, and what else.. As you are mentioning on last post, depends even what solder is used and what temperature is used for specific soldering. It happens to many people to use not good soldering iron & solder and as result amp, make oscilations, heating problems, and the worst scenario, redo all over again.

bekim
 
Hi bekim,
It's exactly the same over here. Same thinking with TV technicians with the same results.

You have a great grasp on the repair industry. Also, your troubles finding parts are the same that I went through. No internet at that time.

-Chris
 
Hi, first post. I just purchased a 2800 that I intend to use with my 3805 to improve (an already decent) sound quality. I have not received the 2800 yet, should be delivered next week. I didn't want to draw any conclusions of it's capability without first ensuring it's reasonable operating condition. You guys appear to have a lot of experience with this model. I was wondering if someone could help me out with some pointers as to what I should be looking for to insure this operates up to spec. Based on this thread, it looks like some (all?) of these will have soldering and/or component issues due to the age. I would appreciate any description of what issues to look for when I open it up. I am obviously not a tech, but have a soldering station and I'm not afraid to use it! Thanks. Blair
 
Hi again and sorry to the OP if I am jumping in on your thread. I was just so glad to see recent talk of the 2800. As it turns out, I received my 2800 today in apparently good condition. The first thing I noticed is the amp is much "faster" then the 3805 alone. It almost appeared as if the music was all playing at a faster tempo. Very effortless dynamics.

Anyhow, my question might be somewhat related to the OP's right channel issue. The one odd thing I noticed when I cycled my 2800 power off and on (while leaving the 3805 on). When first turning the 2800 on, the right speaker pops loudly and plays music immediately. The left speaker does nothing to start with, no music and no pop. The power led then flashes for a few seconds, then the amp clicks, the power led lights solid and the left channel kicks in and begins to play (without any additional pops). I realize that I am new here but if anyone could please help me out and tell me if this is normal I would greatly appreciate it. That toroidal transformer is massive !!

Thanks
Blair
 
It appears the amp is functioning however the right ch. is NOT normal. When you cycle the amp ON, the amp should behave like you described the left ch. The indicator flashes and goes solid after the protection/muting delay and at which point the speaker/protection relay clicks in.

Sounds like the right ch. relay contact is stuck closed (possibly welded together if that channel ever blew up in the past) or has been possibly jumpered for whatever reason. As it stands you have no right ch. speaker protection if the amp decides to take a dump.

I may be wrong but I can only deduce from your posts that you don't have much aptitude in the way of electronics. However if you have a basic multimeter we can do some easy preliminary DC offset measurements at the speaker terminals. Regardless the amp requires some attention.
 
Last edited:
Hi Blair,
Sorry for the delay, I'm not ignoring you.

The output relay may in fact be welded together. The plastic top can snap off easily for inspection, but I'm so sure this is the case, you should just simply replace it. Replacing this relay on any older amp is a good idea, regardless. Look at the solder connections for the relay on the PCB to see if anyone has tried to solve contact problems.

I don't recommend that you attempt to resolder connections yourself. Believe it or not, there is a skill involved and some extra equipment (like a solder sucker - larger one, and solder wick. Solder flux is also a big help here). You will also need to clean up any flux once this work is done. This is really work for a really good audio technician. If you are positive you have the required skills, go ahead. However I would think you already would know what connections to look for. Keep in mind that some bad connections are next to invisible, experience really helps here. Do not just add more solder to the connections!

These are very good amplifiers in the grand scheme of things. There are better and certainly worse products out there. What you have should work great for years and sound better than average. Of course, that depends on your music sources as well.

Don't let personal pride blind you to believe you can handle anything. The more you know, the more you will understand you don't know everything and also that there are many more skills to learn.

-Chris
 
Hi Chris and Bigred, thank you both for your insightful replies. I just got back after being away for the weekend for my son's bday. I have looked and can see a wire "very" crudely soldered inside the unit near the right channel connection. This appears very suspect considering the symptoms and this is obviously not a professionally soldered or wrapped wire so it stood out, and this wire does not appear on the left channel. Assuming the protection circuit was purposely bypassed, I am now concerned that if I were to replace the relay, then I might be facing the original issue which might be more serious then a relay (unless the original problem was only the relay itself)? BTW, I do have a simple (cheap) VOM and basic soldering skills. I had just re-capped my Forte-IIs and originally contacted this forum thinking that I might need to do the same to this amp since the caps in this amp were so old (but maybe originally of higher quality then the Forte).

Anyhow, I have been discussing the issue with the person that just sold it to me. He tried to tell me that it worked fine before he shipped it and it did not have the symptoms that I described and that it must have been damaged in shipping...I now don't think so. I just explained to him about the wire I saw and am waiting to hear back before I touch anything.

Thanks again.
Blair
 
......Assuming the protection circuit was purposely bypassed, I am now concerned that if I were to replace the relay, then I might be facing the original issue which might be more serious then a relay (unless the original problem was only the relay itself)?........
Your concern is very valid but I don't think it's too serious. Having said that, without seeing the amp or knowing what has gone on with it's history nothing can be said with certainty.

As I said, the amp appears to be operating "OK" aside from the protection/relay/muting issue. My experience leads me to believe the original problem was a blown channel and the relay contacts welded together. Now my spidey senses tell me the amp then had "repairs" done by a hacker. A pro certainly wouldn't have left a relay with melted/welded contacts and/or bypassed after repairing everything else.

Obviously any previous repairs/hack jobs now come into question and a determination has to be made. It could be an easy fix or it could become more involved due to "hackeritis". Pictures would be very beneficial here.

While you wait the outcome with the seller, we can at least simply measure the DC offset as an easy preliminary step as I mentioned earlier. Turn the amp ON without any signal or speakers connected and select the "A" speaker switch. Take a DC voltage measurement across the respective speaker terminals. We know the left ch. is low enough not to be tripping the protection and the right ch. who knows but measure both channels anyways.
 
Bigred, I couldn't wait either. A few hours ago, I removed the soldered jumper to see what happened and, of course, no sound came out of the right channel (using speaker "A"). However, while I was in there, I noticed that there is only one speaker signal (one black, one red wire) from the amp to the right channel protection circuit board, but... there are "two" relays per channel. One for speaker "A", the other for "B". So...I hooked up to the right channel using speaker "B", switched the front led to "B" and the right channel worked perfectly, Great !!

Btw, I never tried the "B" connections since I thought it was just directly hard jumped to the "A", but then the fact that there are "A", "B" switches would certainly indicate otherwise.

I looked at both channels amp boards and could not find any differences in component appearence or signs of tampering, like other poorly [re-]soldered components or jumpers that did not belong or new shinny components. As you requested, I just now measured for DC on each channel, but had to use speaker "B". There was no DC on either channel.

Therefore, the status is that [one of] the right channel relays [only for speaker "A"] are permanently open and not welded. Based on this status, could it be simply that the relay failed and not because of any underlying amp issues? Does this POA-2800 model (or the 2200, 2400) have a history of failing protection circuit relays (for no other amp related reason)? I would assume that, if the relay did fail as a result of amp issues, the "B" channel relay would not have been touched if it was not switched on at the time. What do you think?...should I just use it (using "B") and be happy?...or dig further?

Oh, and back to my initial concern from before I received the amp, based solely on the age of this unit, do I have to replace any of the caps (or anything else) to insure it is operating reasonably up to spec? I ask this since I just re-capped my [old] Klipsch Forte-II crossovers based on many forum discussions I read. But then they say that Klipsch used cheap caps, and maybe that does not apply to this amp.

Btw, at one point, I had the right channel (using "B") hooked up through the 2800 and the left channel driven straight from my Denon 3805. This allowed me to clearly hear what the amp improved in the sound quality. More dynamic and fast, good stuff !!

Thanks for your input !!
Blair
 
I was gonna suggest trying the "B" terminals later after the DC offset measurements. You jumped ahead but as I was anticipating anyways, the "B" operates normally on both channels.

I will have to question your DC voltage measurements and/or method because you should have measured some albeit even if some mV's. What scale did you use? The offset is not high enough to cause the protection to kick in but that doesn't necessarily mean it's "ideal". It may take from a volt and often more. Making assumptions when troubleshooting will come back to bite you in the butt but it's probably safe to say the offset and amp are fine. You did mention the LOUD pop initially powering up so I'd still like to get true offset readings.

Anyways, my initial thought on the relay being jumpered or having welded contacts sounds like it ended up being the former. It's your call on whether to just use the "B" terminals or repair the underlying problem on "A".
 
Hi, by the process of elimination, I guess the "A" right channel only has a bad relay but maybe the amp will intermittently go into protection mode for some yet to be experienced scenario. Maybe when I get a more accurate DC reading, this will possibly tell us if something needs to be done with the amp. Since using "B", the right channel no longer pops upon startup.

I only used 10ths of a volt scale since the mV's scale (that I first tried) was just jumping all over the place once I held the plastic leads firmly (and didn't even touch the tips of the leads to anything yet). I guess I would need to buy some better leads to get a decent reading (and VOM). A good excuse to buy more tools, I always wanted one of those Fluke VOM's 🙂
 
Hi, by the process of elimination, I guess the "A" right channel only has a bad relay
Again, you can't make assumptions. It's not necessarily the relay but the drive to pull it closed.

Since using "B", the right channel no longer pops upon startup.
Well of course not, well at least you can't hear it because the relay thus the start up delay are functioning normally for the "B" terminals.


I only used 10ths of a volt scale since the mV's scale (that I first tried) was just jumping all over the place once I held the plastic leads firmly (and didn't even touch the tips of the leads to anything yet). I guess I would need to buy some better leads to get a decent reading (and VOM). A good excuse to buy more tools, I always wanted one of those Fluke VOM's 🙂
A simple VOM with a DC mV scale is adequate for this purpose. Set it for mV's and hook the probs to the terminals. Fluctuation is normal and should stablize reasonably as the temp stabilizes. Leave the amp on say 15 minutes idle and try your measurments. Make sure the speaker selector is switched on for whatever speaker terminals your measuring at.

Again, I'd say the amp aside from the "relay" issue on "A" is operating fine and is healthy. Just use the amp as is on the properly functioning "B" terminals for now and see how the amp operates. You then can make a determination on whether to worry about "A" or not.
 
Hi blair6878,
Btw, is there any place that sells a replacement relay that will fit this? Thanks.
Denon?
It would be much easier to simply get the exact relay and install it. There are so many relays available that you may have a real problem identifying the correct one.

I only used 10ths of a volt scale since the mV's scale (that I first tried) was just jumping all over the place once I held the plastic leads firmly (and didn't even touch the tips of the leads to anything yet).
You have to short the leads together for that test. What you are seeing is random noise, normal operation.

I guess I would need to buy some better leads to get a decent reading (and VOM).
Leads do make a difference here. Any new decent DVM (Digital Volt Meter) will come with good test leads. So you are jumping the gun a little bit here. If the leads youa re using now slip off components and test points easily, buy new (better) test leads for that meter. You don't want to blow anything up.

Both Fluke and Agilent make excellent hand held meters, and Agilent just introduced some really nice meters for around $220 or so. That's one up from the bottom. The extra features in this true RMS reading meter will make your life much easier. Also, the way they read frequency creates a display that updates quickly, faster than a normal frequency meter.

Hi Bigred,
A simple VOM with a DC mV scale is adequate for this purpose.
Well, a simple VOM will not be accurate and may not even have a mV scale to begin with. A simple DVM, or hand held multimeter will be very inaccurate on it's mV scale - if it even has one. Also the shielding in the case may not exist, which will cause very unstable readings. Also, inexpensive digital voltmeters may have such poor accuracy figures that the last digit and a couple counts on the second last digit in the display are meaningless. That's pretty sad, wouldn't you say?

-Chris
 
Chris,

With all due respect, let's not over complicate this situation. His meter does have a mV scale and I stand by my statement that it would be adequate enough for the intended purpose here. I was merely trying to assess if there was any REAL cause for concern with respect to DC on the outputs.

Just as an example, if he measured 20mV's I wouldn't expect a cheap meter to be grossly far off and I'd have no concern. Now if he came back and said I'm measuring 500mV's well um ya. I'm of course all for having high quality and accurate measuring equipment but lets not send him out to buy a Fluke or Agilent for an odd job unless he has money to burn or wants to get real serious. :2c:

BTW, I use my cheap Princess Auto meter for general purpose around the house stuff like working on the boat trailer lights rather than risk wrecking my much more expensive quality Fluke 87V. You know what, cheapy measures well within a mV of the Fluke.
 
Hi Bigred,
With all due respect, let's not over complicate this situation. His meter does have a mV scale and I stand by my statement that it would be adequate enough for the intended purpose here.
Maybe - probably, possibly. For looking at DC offset - yes. For setting bias current, maybe not. After working in a calibration lab, I'm a real cynic. Understand that cheap meters are normally calibrated at 1.9 VDC, not 190mV. Even with the wide tolerances, many meters were not capable of being put into tolerance due to the input range dividers, even when new in the box. The scene becomes more grim as the meter ages. I will say that the mV scale may suffer less than a range that has more resistors involved in the chain.

For DC offset, less than 50 mV is fine, over that and I start looking for why. If he sets bias current, 6 mV may make the difference between crossover distortion and smoking amp. I really wanted to show the difference between the two situations.

Having said this, I also use cheaper meters for non-critical readings as you know. However, I do know where they are reading because, like you, I compared them to something better. BTW, nice meter you have in the 87V. Expensive as well. Most people can afford a used Fluke 23 or 77 though. Those are better choices than a new brand "X" meter. My intention was never to force anyone to buy a new, top of the line meter.

Another odd fact. It's really easy to kill a cheap meter, and yet something like a Fluke 87V will survive all kinds of stupidity. There's a lesson in there somewhere, don't you think?

-Chris 🙂
 
Bigred, thanks for the link to the OEM relay. For $14, I am definitely going to grab one. I just didn't think that they would still be available for an amp this old.

Chris, thanks, I didn't know about shorting the leads. I'm kind of in the dark about most of this stuff. So....some cheap meters display more digits of accuracy then they are capable of measuring. That's just wrong 🙁

Well, I took the plunge and bought a Fluke 87-v and it should be delivered this Tuesday. I know this is way more then I need for most things, but I have actually always wanted a good Fluke and found a used-like-new one for a great price of $175 so I grabbed it. That's OK, I didn't need to eat this month anyway, lol. I also looked at the new Agilent which looked very good as well for $221. The replaceable "skins" were sort of a comical twist. Now lets see if I can figure out how to turn it on. lol. If I understand correctly, even the Fluke 87-v will not be able to adequately check a capacitor for complete functionality...which is one of the things I was concerned about. I'll keep the cheap meter for non-critical things.

Anyhow, the amp is hooked up (to the "B" channel) and working fine for now. I will check for DC and would also be interested in checking/setting the bias when the meter arrives...if you guys feel that is a good thing to do.
Thanks,
Blair