Harman Kardon Citation 17 Pre-Amp One Channel Intermittent

[peterw...@gmail.com]

I have had this beast for about 10 years now, and it has developed a specific symptom - the right channel will cut out entirely - silent-same-as-no-signal on occasion. No pop, no prior acting out, no distortion. Otherwise, no issues.

On the output board is a DPST >NC< 24V reed relay. Meaning (to me) that when it goes into protection-mode, the relay gets energized and output is cut. The schematic (HI-FI Engine) seems to bear this out.

Picture of board here - down the page. http://members.quicknet.nl/gerard.slikker/hkcitation17.htm

There seem to be a number of versions of this relay. Additional part-numbers are: 200-002-7610 and 328-46-B

Questions: Anyone else with this symptom in this device? I would hate to replace the relay only to find that something else is going on and the system is doing its job.

And, may I substitute any decent 24V DPST relay? I have lots of room to play around.

Other notes:

a) It is paired with a Citation 19 power-amp, acquired separately.
b) I have recapped the power-supply, and the output voltages are as they should be.
c) The unit dates c. 1978

Thanks in advance!

Peter Wieck
Melrose Park, PA

[peterw...@gmail.com]

Update: Computer Components is still in business and has ten (10) pieces in stock made in 1978. They are testing them with the intent of sending me two - at a price, of course.

But, I am still wanting to figure out if this is the most likely issue - or should I be looking beyond.

[Dave Platt]

In article <68e28f5e-a05a-44f4...@googlegroups.com>,

pf...@aol.com <peterw...@gmail.com> wrote:

>I have had this beast for about 10 years now, and it has developed a specific symptom - the right channel will cut out entirely - silent-same-as-no-signal on occasion. No pop, no prior acting out, no distortion.
>Otherwise, no issues.
>
>On the output board is a DPST >NC< 24V reed relay. Meaning (to me)
>that when it goes into protection-mode, the relay gets energized and
>output is cut. The schematic (HI-FI Engine) seems to bear this out.

No, it looks to me as if the circuit works the other way around. The
relay is designed to short the output signal to ground *until* the
relay is energized. During normal listening, the relay is energized,
the contacts are open, and the relay "vanishes" from the audio path.

Take a look at how the relay is driven/controlled, from an auxiliary
output on the power supply board.

>Picture of board here - down the page. http://members.quicknet.nl/gerard.slikker/hkcitation17.htm
>
>There seem to be a number of versions of this relay. Additional part-numbers are: 200-002-7610 and 328-46-B
>
>Questions: Anyone else with this symptom in this device? I would hate to replace the relay only to find that something else is going on and the system is doing its job.

Don't have one myself. Here are my thoughts, for what little they may
be worth.

Since the relay is a DPST single-coil, and only one of the two
channels is affected, it seems unlikely that the relay drive is
bad... if it were, both contacts would drop to NC simultaneously and
both output signals would be shorted to ground. So, if it's a fault
in the relay, it must be in the contacts for that one channel -
somehow they're sticking closed. I'd be more likely to suspect a
reed-relay fault where the contacts _don't_ close properly (due
to e.g. lack of contact self-wiping, or the use of a standard reed
relay which requires a wetting current when you'd really want a dry
relay) and that doesn't seem to be the case here.

It _could_ be a case where you've got inadequate current drive to the
relay, and for some reason this affects only one of the two reeds, but
that seems less likely.

I think the first thing I'd suggest, for practical trouble-shooting,
would be to catch the problem when it exists, and then trace the
signal through the output board... see if you've got audio at J1/3 and
then follow it through the output driver and see where it disappears.

>And, may I substitute any decent 24V DPST relay? I have lots of room to play around.

It doesn't look to me as if the protection circuit is at all
sophisticated... the relay is driven from a simple time-delay circuit
which is driven by the +24 regulated supply. The relay driver is an
MPSA13, which is good for 500 mA continuous. The relay itself is just
a short-to-ground NC and the contacts aren't in the normal audio
signal path at all. So, it doesn't look very critical to me... any
small DPST-NC or DPDT 24-volt DC relay would probably serve (just
remember to keep the diode across the coil!)

>b) I have recapped the power-supply, and the output voltages are as
>they should be.

Other possible faults would be a bad 'lytic in the signal path (I've
occasionally seen them go entirely open), or bad switch contacts (input
selector, tape-monitor, etc.), or a broken solder joint that's
undergoing thermal cycling. The "tap the boards with a chopstick"
technique might help pin down the location of the fault.

I think a signal-tracing exercise would be a good idea, before you
replace anything.

[peterw...@gmail.com]

Dave:

You are right on the relay operation - I should have caught that - and it makes sense that a power-drop would also drop output.

I will signal-trace, but given the symptoms, and with proper understanding of the relay operation, I am even more suspicious of the relay. Most of the caps are new (less than 10 years) and the switches, contacts and input jacks are clean. But, testing leads to better understanding - all good.

Thanks for your suggestions!

[John-Del]

Many years ago I saw just one DPST protection relay go bad, but it would cause a noisy output or intermittent crackles as the volume was raised as one set of contacts were pitted.

But otherwise I'd say if one side is good, it's most likely not the relay as Dave pointed out. The best you can hope for is that it quits entirely. Nothing is easier than tracing a mute audio channel when there's a working one right along side of it, and nothing is worse than chasing an intermittent audio channel when it's working..

[peterw...@gmail.com]

On Tuesday, March 20, 2018 at 4:05:27 PM UTC-4, John-Del wrote:

> Many years ago I saw just one DPST protection relay go bad, but it would cause a noisy output or intermittent crackles as the volume was raised as one set of contacts were pitted.
>
> But otherwise I'd say if one side is good, it's most likely not the relay as Dave pointed out. The best you can hope for is that it quits entirely. Nothing is easier than tracing a mute audio channel when there's a working one right along side of it, and nothing is worse than chasing an intermittent audio channel when it's working..

The first test will be that wooden peg on the relay. Agreed that the quiet failure does not lead to the relay first thing.

[Phil Hobbs]

Seems easy to test--work the relay a few times, hit it with a
screwdriver, hair dryer, cold spray....

Cheers

Phil Hobbs

[Ron D.]

I've had bad contacts on relays before. Both audio equipment and a programmable power supply. To fix the power supply, I wrote a program to open/close the relay every 10s overnight and the problem went away.

I'd measure the DC at the speaker terminals before and during a fault first.
I'd expect something initially and probably zero if your loosing contact.

Then do the same, but before the relay.

[John-Del]

He won't be picking up any dc on the outputs. The other channel remains. If it was a DC fault detect, the relay would open and both sides would go out.

Besides, the DC detect is hair trigger on these and the meter probably wouldn't resolve the value before the relay opened. A min/max peak recorder feature on the dmm might catch it but even that isn't guaranteed.

The easiest thing to do if the relay is suspected is to thwack it while the channel is out (but not hard enough to bounce to effect a contact bounce). If the contacts are gimpy, something will be heard on the mute channel.

The problem could be anywhere. If he has pre amp in/out jumpers on the back panel, he could remove them and put in short RCA cables and cross them to see if the drop out moves to the other side. If so, it's before the junction, if not, it's after.

The easiest thing to do is trace it with a scope when the channel acts up. The problem is that once you get near an intermittent circuit with a scope probe, the problem clears up..

[peterw...@gmail.com]

This weekend will be the testing. A new relay (exact fit) is $80 from the same company that made them for HK. Not bad for an obsolete part. but about 10X the cost of a similar unit. But the gentleman at Computer Components stated that their manufacturing specs called for "matched reeds".

The relay is also large enough to open and have a look.

[jurb...@gmail.com]

Check the tape monitor(s) first.

[peterw...@gmail.com]

On Friday, March 23, 2018 at 10:43:08 AM UTC-4, jurb...@gmail.com wrote:
> Check the tape monitor(s) first.

I will.

[jurb...@gmail.com]

It also appears that the hi cut and low cut switches can cause a complete loss of signal.

Remember since there is no big noise there is no DC involved. It also appears that some of the boards are plug in, all of those connections could be bad.

For the switches, worjk them very slowly to find out if in the out or in position they lose it at any time. Also, wiggle the PC boards very slowly and see if you can duplicate the problem that way.

Going fast doesn't get it sometimes. Also when you clean them, after some vigorous action, then go very slowly looking for any cutout or even distortion.

I found an elusive source of intermittent distortion in one channel of a Yamaha receiver by turning the function switch very slowly, and this was after it had supposedly been cleaned already by someone else. And this was not the first time that going very slowly revealed an elusive intermittent. Fast action and brute force are not the best course of action here.

[Chuck]

FYI Yamaha had a whole year of amps and receivers with intermittent
source select switches in the 1990s which couldn't be repaired by
cleaning. They were made by Alps and had to be replaced.

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[jurb...@gmail.com]

>"Yamaha had a whole year of amps and receivers with intermittent
source select switches in the 1990s which couldn't be repaired by
cleaning."

This was a Natural sound type with a silver front and wood case, analog tuner. Had the flat knobs for source and tape select.

The switches were the linear type operated by some sort of cam gizmo from the knob.

I remember on that unit the asked why he had to turn the volume up to like 11 oclock to get the same level as his lower powered Yamaha got at 9 oclock. I had to explain that if they made the linkage on your car so that it was floored when you gave it half a pedal, would it go faster ? That that is a feature, a better control that gives you better control. In this country people often confuse gain with power. The have the volume at 8 oclock and don't realize that some of them were at nearly full power at that point. Extra gain is nice but they took it too far.

[peter wieck]

It is the relay. Tap-on/tap-off.

And, it so turns out that the previous owner had replaced the relay by kluge wiring it on the bottom of the board, holding it down with elephant-snot glue. Which, in turn, was eating the wiring.

In any case, I will order the relay tomorrow and correct the "modifications" made to the board.

[John-Del]

Huh... *I* use elephant-snot glue. It's not as durable as animal hide glue but it's a lot cheaper and doesn't smell nearly as bad...

>A new relay (exact fit) is $80 from the same company that made them for HK. Not bad for an obsolete part. but about 10X the cost of a similar unit.

Is the relay really $80 or were you exaggerating? If it is, replace it with *anything* else that will fit. If you don't have something, I just might.

[peterw...@gmail.com]

On Saturday, March 24, 2018 at 5:18:16 PM UTC-4, John-Del wrote:

> Is the relay really $80 or were you exaggerating? If it is, replace it with *anything* else that will fit. If you don't have something, I just might.

$79.95, no kidding, but including shipping.

a) There is an "anything else" in there now.
b) I get a tested and warranted OEM part from the OEM maker. The warranty includes labor should it be necessary - 10 year.
c) Made with "Matched Reeds" per the OEM spec.
d) And, perhaps the OEM maker will stay in business just a bit longer.

Sure, I can get something that will work from Mouser for less than $12, with shipping. But the reeds will not be matched, and, likely it will be from China, and it will definitely not be original. I dunno, this is not a museum-piece by any stretch, but I can afford it, so why not?

[John-Del]

On Saturday, March 24, 2018 at 6:24:03 PM UTC-4, pf...@aol.com wrote:
> On Saturday, March 24, 2018 at 5:18:16 PM UTC-4, John-Del wrote:
>
> > Is the relay really $80 or were you exaggerating? If it is, replace it with *anything* else that will fit. If you don't have something, I just might.
>
> $79.95, no kidding, but including shipping.


>
> a) There is an "anything else" in there now.

Yes, but I'm sure you could implement another "anything else" solution and have it look fine, work properly, and last indefinitely (elephant-snot glue notwithstanding).

> b) I get a tested and warranted OEM part from the OEM maker. The warranty includes labor should it be necessary - 10 year.

A relay should last 50 years IMO. And while that warranty is impressive, I can't imagine how much trouble it would be to get them to cough up labor costs to replace it. Do they go on your word? Do you have to ship the relay to HK?

> c) Made with "Matched Reeds" per the OEM spec.

Unless I was building a harmonium, I wouldn't worry about matched reeds...[rimshot] I refuse to believe that someone (anyone) could hear an audible difference between audio switched through an $80 matched reed relay and a $5 reed relay from Digi.

[Ralph Mowery]

In article <788f293e-4d61-4088...@googlegroups.com>,
ohg...@gmail.com says...

>
> Unless I was building a harmonium, I wouldn't worry about matched reeds...[rimshot] I refuse to believe that someone (anyone) could hear an audible difference between audio switched through an $80 matched reed relay and a $5 reed relay from Digi.
>
>
>

Don't tell that to the audio nuts. They even go to the process of
replacing the solder in some equipment to a certain type to make the
audio sound better.

They will buy a 6 foot AC cord to go from the equipment to the wall
socket for around $ 100 that is something special. They don't seem to
know there is 25 to 100 feet of inexpensive cable from the socket to the
braker box, and hard telling what from there to the power company
generators.

[peterw...@gmail.com]

Just ship the board to Computer Components. HK is pretty much out of the legacy repair business. CC was their OEM supplier many years ago, and remain functioning to this day. I have to admit that is much of the appeal - how many such businesses remain viable, and how much is their viability, and the jobs they support, worth?

[Dave Platt]

In article <788f293e-4d61-4088...@googlegroups.com>,

John-Del <ohg...@gmail.com> wrote:

>> c) Made with "Matched Reeds" per the OEM spec.
>
>Unless I was building a harmonium, I wouldn't worry about matched reeds...[rimshot] I refuse to believe that someone (anyone) could hear an audible difference between audio switched
>through an $80 matched reed relay and a $5 reed relay from Digi.

Especially considering the fact that in this particular circuit, the
audio doesn't go through the reeds during operation! The reed
contacts _short_ the audio to ground when the power is off, is turning
on, or is turning off. The relay is "normally closed", and opens
(un-shorting the output) during operation once the power supply
stabilizes. When you're listening to music, the reed relay contacts
are out-of-circuit, open, not carrying current, inert, silent, and
out-of-mind :-)

The only impact I can see from having non-matched reeds, would be a
slight difference in the exact timing of the un-muting (at power-on)
and re-muting (at power-off). Somehow I can't see a millisecond or
two of variation at those times, as being an issue.

[peterw...@gmail.com]

Hmmm... In the Computer Components Catalog, this relay is between several mercury devices, but not marked as such. Might explain the price.

[jurb...@gmail.com]

>"Especially considering the fact that in this particular circuit, the
audio doesn't go through the reeds during operation! "

Some companies prey on audiophoolery. Ten buck peer foot wires, CD demagnetizers and all that.

When are the going to come out with a DVD rewinder ? I mean, for the guy who has everything...

[tabb...@gmail.com]

done years ago. Google.
Why it was done I don't know, I hope as a joke.


NT

[John-Del]

Yeah, they were jokes.

More disconcerting are the fools that "saw" improvement in television images by use of an external power conditioner. While the conditioner made have had some utility as a surge protector, they had no effect on the picture, which is why most fools bought them.

[peterw...@gmail.com]

The relay arrived yesterday. About 1" x 1.25" x 3/8" and weighs as much as two quarters and a dime (coins). Six pins. And, it has the original Harman-Kardon part number on it as well as the CCI part number.

[peter wieck]

With the new relay in place, the unit is now operating properly and reliably. What I took out, once the elephant snot was removed, was a 12V sub-miniature US-made relay on a 24V system. That it lasted 10 years is remarkable. I Looked it up. No longer made and the company no longer listed.

[Peter W.]

Follow up from 2.5 years ago. Over the last few days, I did my annual-re-making-of-all-connections exercise on the HK system (Citation 17/18/19, AR M5 speakers, Revox turntable, Revox cassette player, Sony CD Changer, ). Four things arose:
a) The phono-input (17) was fuzzy. Fix was re-solder the right-channel ground on the input jack - HK components of that era are notorious for cold-solders. Luckily it was visible.
b) Minimal hum, both channels, only when the 17 was ON. Cleaned up two internal grounds, went away.
c) The *other* channel became intermittent - but full output from the 17, so, within the 19. Much longer fix - with the wooden stick method, literally poking around. Found the 4-section Molex connector (only two sections connected) was intermittent at the ground. NOT the connector, but the square pin through the driver board was cold-soldered and/or had cracked at the board. Interesting inasmuch as it tested for continuity, but was intermittent in operation Re-soldered - good now. I also re-soldered all the pins on the other Molex connectors - figuring they were all done at the same time, likely by the same individual with the same equipment.
d) The AR woofers both needed new surrounds. Which by pure blind luck, I had sitting around with the proper glue. Took about 4 hours each - mostly as that is how long it takes the glue to set. Better now!


While I was there, I adjusted the muting threshold and stereo thresholds for the 18. The target stations have reduced power slightly, making the adjustment more of a convenience than a necessity.

Anyway, no more troubles. Next: the main system.