King/Cello Tape Stage Coupling Caps

[John Chapman]

Hello!

I have been still very much enjoying both tape and the King/Cello tape stage. Last week I finally dove inside the box to have a look around. I am a bit embarrassed that I did not open it sooner - on the gear I build and sell I always wanted to put stickers on that read "Warranty Void If This Seal Is NOT Broken" to encourage folks to at least have a peek inside and maybe play a bit with mods, etc.

I popped the top intending to scope out the coupling cap sizes as I had read of folks swapping caps as a mod. Tracing it out is seemed there were 3 coupling caps in parallel at the input and none at the output. The 3 input caps were a 470uF electrolytic and a small film bypass on that along with a 1uF film. Here is the exchange with Charlie slightly edited for clarity and posted with his permission:

From Charlie:
"
The "mid-sized" paralleled cap - probably around 1 ufd is the one I "play with" - of course you can do whatever you like. I put in as large a value I can consistent with it "fitting under the hood". I have fitted close to a 1uf gold/silver/oil Mundorf - haven't tried a Duelund or Clarity or whatever else cap yet - too expensive - but you are welcome to do that and PLEASE REPORT your findings.
"

From me:
"
Thanks for the info. Will play and report back. I have to ask if we need anything from the tape heads? Would there be any DC to block or any other reason to use one there circuit wise? Is there a potential DC offset at the input to the tape stage that could be a problem for the heads maybe? Will report back after I play in any case.
"

From Charlie:
"
Cello started with no input cap but added one later on. With a scope I observed a "switching on" voltage transient which while not severe to damage anything could magnetize the head after a while so i didn't think it prudent to eliminate it. You could try it just demag the heads - or leave the unit on
"

From me:
"
Well I gave myself the day off yesterday or the afternoon off anyway. I had been meaning for so long to dive inside the tape stage but other stuff always seems to fill up the days..... Here are a few notes on stage 1 of the playing:

1- After your later e-mails I thought I may get away with no cap. A quick check revealed that we have about 150mV of dc offset at the input to the circuit. Although the turn on transient you mention could be handled with a delay switch or relay, etc this constant offset would be a problem for the heads I think. It is higher than I ever would want to run our attenuator autoformers for instance. It could be possibly tuned out by playing with supply rail voltage matching but that would be a bit tenuous as it could drift over time so a cap (or some way to isolate the heads) is indeed needed.

2- I put the unit on the Audio Precision here. It was fun to see the tape curve and the phono curve. Tape is a much simpler correction curve - I am sure you guys know this but it was interesting to see. Both channels match very well and THD was very low. It was 0.09% typically but with such small input voltages (1 to 2 mV depending on what I was looking at) that is very good as the noise floor is creeping in big time with only those small signal levels at play.

3- First thing I did was pull the 470uF electrolytic cap and it's small bypass cap on one side. The Cello guys would run circles around me design wise but I have not used polar electrolytics as coupling caps like that (could be that 150mV DC offset is there to bias that cap?) so I was curious how removing it would change things. Also I have always preferred a single decent cap to a nest of various values. Popped it back on the test gear and both sides (with and without that cap) still measured identical right down to 10Hz. This was good. I was specifically watching for low frequency rolloff when reducing that coupling cap but saw none.

4- Next I swapped a 0.47 uF V-cap in that side. These were the only decent caps I had handy. Still measured identical. Good again. Swapped out the other side and all still fine.

I installed it back in the system and one tape later (edit - now several tapes later) it is sounding very good. I am afraid I am the worst guy in the world at making sonic comparisons so I won't even try to compare to stock. Years back I spent a great deal of time making comparisons while designing gear and it burned out that section of my brain..... I will leave it like this for a time and then get used to it before swapping in some mundorfs and maybe the new copper V-caps but will do my best to report on sonics at that point. I am left curious just how low we could go in value - next time I have it out I will play with some smaller values to check that affect that'll have on the curves.
"

End of e-mail exchange.

Anyway - I thought it might be of interest and was also curious what others have found when swapping these coupling caps.

Thansk!

John

[Steve Williams]

Nice bit of info John. Thanks

[RogerD]

Hello!

I have been still very much enjoying both tape and the King/Cello tape stage. Last week I finally dove inside the box to have a look around. I am a bit embarrassed that I did not open it sooner - on the gear I build and sell I always wanted to put stickers on that read "Warranty Void If This Seal Is NOT Broken" to encourage folks to at least have a peek inside and maybe play a bit with mods, etc.

I popped the top intending to scope out the coupling cap sizes as I had read of folks swapping caps as a mod. Tracing it out is seemed there were 3 coupling caps in parallel at the input and none at the output. The 3 input caps were a 470uF electrolytic and a small film bypass on that along with a 1uF film. Here is the exchange with Charlie slightly edited for clarity and posted with his permission:

From Charlie:
"
The "mid-sized" paralleled cap - probably around 1 ufd is the one I "play with" - of course you can do whatever you like. I put in as large a value I can consistent with it "fitting under the hood". I have fitted close to a 1uf gold/silver/oil Mundorf - haven't tried a Duelund or Clarity or whatever else cap yet - too expensive - but you are welcome to do that and PLEASE REPORT your findings.
"

From me:
"
Thanks for the info. Will play and report back. I have to ask if we need anything from the tape heads? Would there be any DC to block or any other reason to use one there circuit wise? Is there a potential DC offset at the input to the tape stage that could be a problem for the heads maybe? Will report back after I play in any case.
"

From Charlie:
"
Cello started with no input cap but added one later on. With a scope I observed a "switching on" voltage transient which while not severe to damage anything could magnetize the head after a while so i didn't think it prudent to eliminate it. You could try it just demag the heads - or leave the unit on
"

From me:
"
Well I gave myself the day off yesterday or the afternoon off anyway. I had been meaning for so long to dive inside the tape stage but other stuff always seems to fill up the days..... Here are a few notes on stage 1 of the playing:

1- After your later e-mails I thought I may get away with no cap. A quick check revealed that we have about 150mV of dc offset at the input to the circuit. Although the turn on transient you mention could be handled with a delay switch or relay, etc this constant offset would be a problem for the heads I think. It is higher than I ever would want to run our attenuator autoformers for instance. It could be possibly tuned out by playing with supply rail voltage matching but that would be a bit tenuous as it could drift over time so a cap (or some way to isolate the heads) is indeed needed.

2- I put the unit on the Audio Precision here. It was fun to see the tape curve and the phono curve. Tape is a much simpler correction curve - I am sure you guys know this but it was interesting to see. Both channels match very well and THD was very low. It was 0.09% typically but with such small input voltages (1 to 2 mV depending on what I was looking at) that is very good as the noise floor is creeping in big time with only those small signal levels at play.

3- First thing I did was pull the 470uF electrolytic cap and it's small bypass cap on one side. The Cello guys would run circles around me design wise but I have not used polar electrolytics as coupling caps like that (could be that 150mV DC offset is there to bias that cap?) so I was curious how removing it would change things. Also I have always preferred a single decent cap to a nest of various values. Popped it back on the test gear and both sides (with and without that cap) still measured identical right down to 10Hz. This was good. I was specifically watching for low frequency rolloff when reducing that coupling cap but saw none.

4- Next I swapped a 0.47 uF V-cap in that side. These were the only decent caps I had handy. Still measured identical. Good again. Swapped out the other side and all still fine.

I installed it back in the system and one tape later (edit - now several tapes later) it is sounding very good. I am afraid I am the worst guy in the world at making sonic comparisons so I won't even try to compare to stock. Years back I spent a great deal of time making comparisons while designing gear and it burned out that section of my brain..... I will leave it like this for a time and then get used to it before swapping in some mundorfs and maybe the new copper V-caps but will do my best to report on sonics at that point. I am left curious just how low we could go in value - next time I have it out I will play with some smaller values to check that affect that'll have on the curves.
"

End of e-mail exchange.

Anyway - I thought it might be of interest and was also curious what others have found when swapping these coupling caps.

Thansk!

John

I would swap those 470uf caps depending on the voltage with a blackgate version,maybe the 470uf 100v ,they bring 100.00 a piece now,but the Cello is 5000.00 so it's all relative.

Also, that 150mv dc offset number seems very high to me and could degrade the sound I would think. Is there a reason for that in the circuit design?

"I put the unit on the Audio Precision here. It was fun to see the tape curve and the phono curve. Tape is a much simpler correction curve - I am sure you guys know this but it was interesting to see. Both channels match very well and THD was very low. It was 0.09% typically but with such small input voltages (1 to 2 mV depending on what I was looking at) that is very good as the noise floor is creeping in big time with only those small signal levels at play."

That 1 to 2 mV is where it should be.

[John Chapman]

Roger,

From my testing so far I can't see why we need that 470uF in there at all. I like the simplicity of just a single coupling cap rather than the nest of them and unless I am missing something (quite possible....) then I will just leave it out all together. The frequency response curve was dead on the same with and without it in place.

My guess was that the 150mV offset might have been intentional to bias the polarized electrolytic cap. If there were no offset then there would be no need for any cap at all and if that were the case I would be hard pressed to see why you'd want to use one. I should clarify as well that the 150mV offset is NOT at the input jacks and it poses no danger to the tape heads - the input cap blocks that offset and isolates the heads from it. It is only seen after the coupling cap at the input to the gain stages.

Finally, yes - I was using 1 or 2 mV into the unit to mimic the tape head output (or phono when I was looking at that). That is what I do when testing phono stages here as well. Much more than that could drive the output to it's max signal level on the 50db or 60db gain settings so I was using 40db of gain for all these tests. Actually in my system I run at the 40db gain setting with the tape input all the time as it sits in a nice zone gain wise there.

Thansk!

John

[Steve Williams]

John

Most of us (recently at least) have had their head amp made without output coupling capacitors

[microstrip]

John,

Are you measuring the 150mV with the head connected with or without the capacitor? This offset is most probably due to the current (few microamps) of the base of the input transistor going through the input resistor and should be harmless. Only at switch-on and off you can experience problems of excessive current through the head.

Anyway, if you are using an electrolytic capacitor this offset is very welcome!

[John Chapman]

Hello!

Steve: Yes - my build also has no output coupling caps. It is the input coupling cap(s) I am playing with.

microstrip: The 150mV is measured on the circuit side of the input coupling caps. It is with no head connected but it is AFTER the input coupling cap - ie the cap isolates this offset from the heads and as a result protects it. I think I had the input shorted at the jack when measuring - usually do that but can't recall for sure if I did in this case. It could be that with no coupling cap in place and the head connected that the offset would indeed get low enough to be no issue. I will test this some more next time I have it open - but that will be a few weeks off with my schedule this next couple weeks..... Maybe someone else will beat me to it.

I would love to not need either input or output caps in the signal path but I want to get a better handle on it before eliminating it. Even a small DC offset over time would tend to magnetize the heads so I would only eliminate it if there was no chance of this. The turn on / off transients are not such a concern since they could be handled in other ways.

Thansk!

John

[stellavox]

Here's part of the exchange that John left out:

"I understand your concern about paralleled caps however as you know finding a single one that sounds good, will go down to low frequency AND fit inside are mutually exclusive criteria. Just finished reading both (Bob) Cordell's new power amplifier book and (Douglas) Self's new low-level design book (BTW, both are HIGHLY recommended and fairly easy "reads"). Both Cordell and Self did a lot of measurements on electrolytics; Cordell found that increasing the size of the electrolytic to two or more times the "desired calculated value" reduced measured distortion to the background level of his test equipment, while Self seemed to come down on the side of using BI-polar electrolytics. SO my current (new) tactic is to use "large" bi-polars - new enough that I haven't tried it on any tape pre's (in place of the 470mf polarized), but your measurements seem to back up the large value theory. More to experiment with!"

Oh and Roger D, the price is (still) $3750.

Built up a new battery-powered supply for the "purists" in the room.

Also, finally wired up the quarter-track head in my MCI playback machine so I could also listen to some of those tapes that have been stored for years. WOW - GREAT listening in that format also!

Cheers - and THANKS John

Charles

[MylesBAstor]

Here's part of the exchange that John left out:

"I understand your concern about paralleled caps however as you know finding a single one that sounds good, will go down to low frequency AND fit inside are mutually exclusive criteria. Just finished reading both (Bob) Cordell's new power amplifier book and (Douglas) Self's new low-level design book (BTW, both are HIGHLY recommended and fairly easy "reads"). Both Cordell and Self did a lot of measurements on electrolytics; Cordell found that increasing the size of the electrolytic to two or more times the "desired calculated value" reduced measured distortion to the background level of his test equipment, while Self seemed to come down on the side of using BI-polar electrolytics. SO my current (new) tactic is to use "large" bi-polars - new enough that I haven't tried it on any tape pre's (in place of the 470mf polarized), but your measurements seem to back up the large value theory. More to experiment with!"

Oh and Roger D, the price is (still) $3750.

Built up a new battery-powered supply for the "purists" in the room.

Also, finally wired up the quarter-track head in my MCI playback machine so I could also listen to some of those tapes that have been stored for years. WOW - GREAT listening in that format also!

Cheers - and THANKS John

Charles

Try stuffing these into your tape preamp...or stuffing your preamp into these caps. Big MF's

[John Chapman]

Hello!

Thanks for posting that info - I did not actually get that section of e-mail exchange here and just scoured my in-box to double check.... My testing so far did not show any sign of low frequency rolloff without that 470uF cap and that is what I expected to see when I yanked it out. Once I get time for another round of checking I will dive in and do some more testing then report back.

I have some 10uF dueland caps around here - I used them to protect a line of ribbon tweeters on a line array here - and although they won't fit inside they would be fun to try anyway. The ones I have are flat copper foil types as I recall and not as wild looking as those ones Myles posted!

BTW - since this coupling cap is at the input to the stage (first thing the signal rides through after the jacks) space is not necessarily an issue - we could locate that coupling cap outside the box at the input to the tape stage and simply wire it in series with the cable from the tape heads. Bit silly but it would not be the first time a diy audio experiment got silly.

Thansk!

John