Look for help reviving Foreplay

[slomatt]

After a bit of a break I've been sucked back into the DIY audio world and am very excited to see the upgraded Bottlehead forum.

I built a Foreplay (I, II?) sometime around 2001 and used it in my system for quite a long time with a NAD 314 providing amplification. There was always some hum, but not a offensive amount. Then about 2 years ago I decided it would be a good idea to upgrade to gold RCA jacks and the hum suddenly went crazy. The only difference I can think of is that these new RCAs are insulated, so currently the star ground is decoupled from the chassis plate.

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fwww.blankwhitepage.com%2Fgallery%2Falbums%2Fforeplay_preamp%2FPreamp.jpg&hash=9b580c140b8af60705bab54f15747bba3dad9341)

Additional pictures at: http://www.blankwhitepage.com/gallery/foreplay_preamp

The above pictures are a bit old, here are the current specifications:
- Foreplay from around 2001. I think it is the v1, but could be the v2.
- Anticipation C4S upgrade.
- Pretty sure I added a filament snubber (it was a while ago)
- Insulated RCA jacks
- All silver internal wiring (aka. it was a pain to solder)
- Changed to a "mostly" star ground layout.

With my renewed interest in DIY audio I'd really like to get the Foreplay back into working order and deal with the hum issue. Here's the evidence I have to date.

- Hum is present on both channels and sounds like A/C hum.
- Tubes are significantly microphonic
- Today I noticed that the hum on the right channel decreases significantly in a ~5 degree area of the volume control potentiometer. This seems very strange, the hum is extremely loud, then as the pot is tuned it suddenly reduces and then comes back as the sweeper moves just a short distance.
- Other than the above the hum does not change based on volume setting or input selection.
- I've gone through and checked all the grounds for cold joints.

Here are some things I'm going to try to fix the problem, I'd really appreciate feedback on these and any other suggestions the community may have.

- Tie the DC star ground to the chassis ground. The hum got worse when I installed the insulated RCAs which had the effect of isolating the chassis ground and the DC ground.
- Do additional experimentation with the volume pot to see why the hum reduces when the sweeper is in a ~2mm range.
- Convert to DC heaters. I'm planning to rectify the transformer output, my concern is that this might increase the voltage slightly. Does the polarity of DC voltage to the tube matter?

If none of the above work my next option is to rebuild the pre-amp from scratch, but I'd really like to avoid that step.

Any help or ideas is greatly appreciated. Thanks!

- Matt

[Paul Joppa]

You might search the old forum, but IIRC the power transformer can't support rectified heater power - it will overheat. That's one of the reasons the FP-III has a relatively HUGE transformer ...  most people found wall-wart type DC heater supplies.

You might try some other 12AU7s, microphonics vary a lot from tube to tube (not just manufacturer to manufacturer).

Another possibility is a gain budget problem - a sensitive amp/speaker combo will pick up hum and noise. See the white paper on signals and noise on the "Community" page.

The narrow range of hum reduction is just plain bizarre. All I can think of is two hum sources which happen to be out of phase and when the relative levels are adjusted just right they cancel. That's not much help - means you might have (at least) TWO problems to track down. But start with the central ground point, was that T13/T14?

[Grainger49]

Matt,

That is a lot of good information.  If I have read it right you have two problems - hum and microphony.  I have a FP 2, yours looks like a FP 1.  They are much the same except that the heater voltage is different.

Microphonics are tube related.  I have never had the problem and my circuit is nearly identical with yours.  Change the tubes you will most likely get rid of the problem.

Hum is the proverbial can of worms.  Two things that you posted ring up with me.  The "silver wire" that was "a pain to solder" could be the long grain Magnet wire that Doc sold as an add on throughout the life of the early Foreplays.  It was coated and the insulation had to be burned off.  This could be the source of your problem.  The reason I am guessing this is I have used silver wire and it is pretty easy to solder, it just sinks the heat from the iron and takes a little longer to heat.

The second thing that rang up in my brain was that you changed the RCA jacks.  The better jacks have two insulators (one on top of and one below the plate) and a solder tab that should be grounded.  The standard way was to run a wire (see magnet wire) through the lugs and finally to the central grounding point.  You have described star grounding so I'm guessing you ran a wire to each and to ground?  Check the resistance of each RCA jack outer conductor to the chassis or to the circuit common terminal (T14).  It should read very, very close to what you read when touching the meter probes together.  That is you should get the same or a little higher.  Over an ohm is too much.
You should try is jumping Terminal 13 to Terminal 14.  That is, tie the circuit common to the chassis if you haven't already.  This often eliminates hum, very rarely increases it, thus the temporary jumper first.  And you have that planned.

I'm totally at a loss about the hum and volume control decreasing it at a certain point.  That is a strange one.

The DC heaters help many.  I am using some very sensitive speakers right now, Altec 846Bs, and have AC heaters, there is no hum.  But the DC heaters are something to try.

The "Snubber" circuit is probably on the high voltage.  That is what I remember as VoltSecond's RRSF Circuit.

Post your results.  Your beautiful Foreplay can be incredibly quiet.  From the looks of your build, you are skilled and success should be near.

P.S. I told Doc/David this could be moved to Legacy Products since they don't build it anymore.

Edit: I added below a picture of my FP 2 pretty early on.  It shows the tinned copper wire that connects the Input RCAs, Output RCAs, Volume pots carry the cathode resistors to ground and finally to the circuit common, which you can see is T14.  T14 is jumpered to the chassis at T13.

[slomatt]

First off, thank you for the replies and for moving this to the proper forum.

This morning I shorted the input for the right channel and measured 0.5mV AC on the output of that channel. Unfortunately I don't have a scope so I can't provide more accurate data, but I'm assuming this is the hum I hear. What is strange is that the left fluctuates from .2-.8v on the output! This is obviously huge, and strange since I used this amplifier for ~6 years before I put it in storage for a year, the huge voltage on the left channel definitely wasn't there before.

Paul,

Amplifier gain is definitely part of the issue, I use this preamp with both a NAD 314 amplifier stage and a "gainclone" chipamp I've built. With both amplifiers I only need the first 1/4 to 1/2 turn of the Foreplay's volume knob. For speakers I've tested with an 8ohm ~88db/1m sensitive speaker and also a 4ohm ~01db/1m speaker, obviously neither of these would be considered "high efficiency".

I'm very confused by the narrow range of hum reduction, my first thought was some kind of mechanical effect such as a loosly mounted pot, but everything is very solid.

T14 is the star ground for the DC side, T13 is the chassis ground. This morning I connected T13/T14 and didn't see any measurable difference in the hum, tonight I'll hook up to an amp and see if I can hear a difference.

Grainger 49,

I've questioned these tubes since day 1, I agree with you and Paul that trying out a new set is definitely in the future. Can anybody recommend a reliable tube vendor?

The silver wire I used is actually 4 nines (99.99%+ pure) solid silver wire. I got fancy and also used very high silver content solder, which requires a lot of heat to get a good joint. For 5 years I had acceptable hum levels with this amp, it was only after the new RCAs the hum increased drastically. I've checked all of the RCA solder joints and couldn't find anything.

My RCA setup is exactly as you described. The jacks have a insulator on the top and the bottom and the ground connects through the lugs. As per the original design this ground then runs through another connection point and then to the central ground. I'm wondering if hooking all of the RCAs directly to the star ground would help.

In your email you say to check the resistance of each RCA to the chassis. Currently the chassis is electrically isolated from the DC ground, this is due to the insulated RCAs. Previously the chassis and DC ground were connected via the older non-insulated RCAs. As I mentioned this morning I connected the DC ground and the chassis ground and didn't see a measurable decrease in hum, I'll test more this evening.

I think you are are exactly right about the HV snubber, I can't find my notes but I'm pretty sure it's VoltSecond's circuit which I added on right after I built the preamp.

It is possible that my AC heater wires need to be moved. They are twisted together, but at a point they do get fairly close to a signal wire or two (.25" or so). I'll likely re-run those, only problem is that with the C4S it is REALLY crowded in there.


Thank you again to both of you for the input. At the first Burning Amp event this preamp was used in the side room and received many very posiitive comments, I'd really like to get it back into working order.

- Matt
 

[Grainger49]

Matt,

Looking at those readings there is something wrong.  The difference between the right and left channels would rule out the power supply.  Often moving a DIY project can loosen solder joints that have worked for years.  Is the hum the same on all the inputs?

I edited the post after the email was sent to say measure to circuit common, which is T14.  Since you have swapped out the RCA jacks I would check both the center conductor and grounds on the left channels.  It is best to measure from the jack itself to the destination.  On the output jack that is the output cap for the hot but the inputs hots terminate on the selector switch.  If that has been done reheat all the solder joints.  I would check all the ground/circuit common solder joints.  That is a frequent source of hum.

And, again, yup, I've used 4% silver solder, nice sounding but a PITA to solder.

About using the star ground for the RCA jacks, I don't have that on mine and have no hum.  I've used it extensively with a ST-70, Krell 250, Chipamp GainClone and my Paramours.  But each case is different.  

The heater wires look good in your pictures.  They are twisted tightly and look like they are close to the top plate.  They also look to cross the audio path wires at 90 degrees.  That should keep the AC out.

[Paul Joppa]

Re-reading this thread, I don't see any mention of swapping tubes. If the noisy tube is really bad it will show up that way pretty quick(!); sometimes swapping tubes just cleans the pins enough to solve a problem.

Another common problem is broken grid stoppers - the 220 ohm carbon composition resistor at the grid pins (2 and 7). A floating grid can pick up electric field noise very easily and still appear to function.

As always, checking the resistances and voltages on the tube pins is a good thing to do when you are stumped.

[slomatt]

Just a quick update, due to the holidays I haven't had time yet to do more debugging. Hopefully I can find some time tomorrow morning.

- Matt

[slomatt]

This morning I discovered that my shorting RCA jack on the left channel was broken, once repaired I now measure ~3mV of hum on each channel. I also tested the resistance from the common ground T14 to the RCA jacks and the pots, each is within 0.3ohm of what my meter reads when the leads are shorted.

Next up I'll go through the manual and re-check all internal voltages. Oh, and based on my notes I used the Radio Shack 97/3 copper solider recommended in the manual.

I've never been really happy with the dressing of the heater wires, and based on pictures of other builds the turns-per-inch of my twisting might be insufficient for good cancellation. Another possible issue is that the heater wires between the tubes are "floating" about 3/4" above the chassis, I'd like to re-run them down as close to the chassis as possible. Is it a problem if the twisted heater wires are very close to a signal ground?

Here's an up-to-date photo of the bottom of the preamp. In it you can see the ground buss for the RCAs, the HV snubber, and the star ground on T14 used by the DC part of the power supply. You can barely see the red/white twisted heater wires between the two tube sockets.

(https://forum.bottlehead.com/proxy.php?request=http%3A%2F%2Fwww.blankwhitepage.com%2Fgallery%2Falbums%2Fforeplay_preamp%2FIMG_2738.jpg&hash=abafa84633acd021cc9fc473a0583a14e7c6fabd)

I'll have time to do more debugging in early January. Thanks again for the help, and happy holidays everyone!

- Matt

[Doc B.]

Just to clarify, you had measured 0.5mV rms on one channel and 0.2 to 0.8mV rms on the other channel before fixing the shorting jack, and now you are getting 3mV rms on each channel? Or is it 0.3mV rms? 0.3mV rms of hum would be within spec for a Foreplay 1 or 2.

[Grainger49]

Matt,

I re-read your posts again.  Beyond Doc's question about the decimal point on hum, I wondered if you have noise in the system still?

Getting the heater wires close to the top plate may help.  But first you can get a wall wart from RS and bring DC in to see if the noise is heaters or grounding.  Looking closely at your picture I see that the heater wires come all the way from the transformer to the B channel tube socket.  It would be messy replacing them just to check.  Maybe not.

The real question first should be do you still have the noise as described that changes with the volume pot?  If so, maybe it is heater wiring.  The preamp was silent for quite a while.  The only change was the RCA jacks.  I'm still suspecting them.

[slomatt]

I hope everybody had a nice holidays and a happy new year.

This weekend I finally had time to do more debugging on the Foreplay. I completely re-ran the AC heater wires with a tighter twist and improved wire layout. I also re-flowed all ground connections.

Tonight I checked the resistances and voltages and confirmed that all conform to the values in the assembly manual.

With the inputs shorted I measured 0.3mv(rms) on the right channel output and 76.1mv(rms) on the left channel output. Based on Doc's post above the right channel is within the expected range, but it seems that the left channel voltage is out of spec. When I connected the outputs into another preamp that I use as a "master volume" the audible level of hum seemed about consistent between the two channels, but with an CD player connected to the input it immediately became apparent that there is an issue.

With the Foreplay's left channel volume set at the lowest setting I still get enough signal into the other preamp that there is audible noise even with both volumes all the way down. After mucking around a bit I discovered that when I push down on the left volume knob the sound disappears as would be expected with the volume at the lowest setting, releasing the pressure causes the sound to return. Further, if I flex the chassis plate at all it has the same effect.

I hooked a multimeter between the signal ground (T14) and the center pin of the volume pot which is the signal feed to the tube. On the right channel this yields around 2 ohms which indicates that the input is shorted to ground resulting in no input signal. On the left channel I measured 40.9k ohms with the plate not flexed and the expected 2 ohms with the plate flexed. Based on this my assumption was that flexing the plate was causing a mechanical short, or exposing a bad connection, or that something is wrong with the pot. I've re-flowed all three connections on the pot and "chopsticked" the wires, in doing so I determined that if I push down on the pot terminals with the chopstick the resistance drops to the expected 2ohms.

So, is it possible that the wiper in the pot is broken and no longer touching the coil and that when I "flex" the chassis it moves the wiper onto the coil which grounds the input signal and results in the lowest volume setting? With the chassis not flexed the pot is at 40.9k which corresponds to a "loud" setting and is probably causing the 76mv of noise on the left channel.

Very strange... though this would explain why the hum seemed to disappear when the volume pot was in a short range, I must have been applying enough pressure to make it work properly.

Does this theory make sense? Can anybody else think of other possible causes? Is a sweet whisper in my future?

Thanks again for the input.

- Matt

 

[Grainger49]

Matt,

Excellent information!  You have done anything that we would have asked you to do.  So now I think you have isolated the problem to the pot.  There are a lot of choices out there.  You can buy a replacement, I would buy two.  You are lucky that the pots are so easy to get to.  Probably the only thing in the circuit that is so open.  The FP 2 has a very short signal path which results in the tight layout.

Mouser has many replacements.  I think that there were discussions that you can use 50k pots which will give you a greater selection.  You could put in a volume and balance.  That takes some serious sitting down and thinking about the wiring.  A balance control doesn't work the way you think it does.  I have installed a few without instructions and I did a lot of head scratching along the way.

And, just because, here is a link I got from the old board.  I wish I could remember who to attribute it to:

http://www.geofex.com/Article_Folders/potsecrets/potscret.htm

I found it interesting.

[Paul Joppa]

In addition to Grainger's suggestions, the PEC pots are highly regarded; DigiKey has them - search for Precision Electronic Components. You would want the MIL-spec 2 watt KA series (logarithmic taper) which are I think exact replacements. These are sealed to keep dirt out, and the wiper is a carbon puck matching the resistance element for reliable contact.

And of course the Sweetest Whispers is still around as a kit, now in the form of a shunt-mode control. It offers some superior connection modes but would take a change in wiring.

Gold Point among others make quality high-resolution switched volume controls. They tend to be expensive, an a par with the Foreplay kit cost just for attenuators.

[slomatt]

Grainger, Paul,

As always thank you very much for the helpful discussion.

After reading Grainger's post this morning I ended up sketching out some volume/balance designs while riding the BART (publish transit this morning). It's definitely a tricky problem, I ended up with a solution that uses two dual pots but with the downside that instead of the balance progressively muting one channel while not effecting the other instead it would decrease the signal on one channel while actually increasing the signal on the other. This is not the desired effect.

It's somewhat of a moot point since I'm fine with dual pots or a stepped attenuator, which leads me to the two options I'm considering:

A. I already have a multi-position dual gang rotary switch sitting around so I would "just" need resistors to build a stepped attenuator.

B. The pots that Paul recommended sound like a good candidate.
Here's a 100k Logarithmic 2W for $11.50 each.
    http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=KA1041S28-ND
And here's the 50k version.
    http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=KA5031S28-ND
There are also dual version of each, but based on the 10% tolerance I'd expect quite a bit in deviation between the two channels.

What are the advantages/disadvantages of using a 50k pot instead of a 100k pot? The ones that come to mind are that it changes the input impedance, and if there is some capacitance in the tube it could change an RC value. Otherwise since the pot is setup as a voltage divider I would think the volume range would be the same, assuming a high input impedance on the tube. Now, I have to admit I don't know much about tubes and I took analog electronics a while ago so please correct me if this is wrong.

One issue I currently have with the Foreplay is that it provides much more gain than I need with my solid state amps and I usually only use the first 1/4 to 1/2 of the volume knob. Is there an easy way to adjust the gain of the Foreplay without decreasing sound quality? This could also help to reduce the ~0.3mv of hum I'm still seeing.

Thanks again for the assistance, I'm glad that I revived this project and am making progress.

- Matt

[Paul Joppa]

Your analysis is spot on at every point, as far as I can tell.

You can attenuate the input by using a series resistor - for example, a 50K from selector switch to the "top" of a 50K pot will reduce the input level to half the voltage, i.e. 6dB. A 90K resistor with a 10K pot is 20dB, a factor of 10 in voltage. If you put the series resistor between the RCA jack and the selector, you can use a different attenuation for each input - that's how the FP-III is set up.

However that will not reduce the preamp's internally generated hum and noise. If the system has too much gain and the preamp noise is audible, an attenuator at the power amp input would be more useful. For details, see my white paper on signal to noise ratio, on the Bottlehead Community page.