K-502 Modifications (capacitors, etc.)?

[Spraiski]

I am looking at ordering a K-502 as a first tube amp project.

In the spirit of building the best amp that I can (within a limited budget) I have been looking at the various suggestions of upgraded capacitors, addition of snubber capacitors, etc.

The problem is that I am not familiar with the various types of capacitors that I might want to use (I understand that there are a number of different options with type of capacitors, brands/manufacturers, etc). As an example of this, I see 'Orange Drop' capacitors listed on the AES website, but I don't know which series of capacitor I want?

As well, I am not sure of the values that I need/want as I have not yet ordered the K-502, and do not have a schematic yet.

Can anyone suggest specific capacitors from AES or digikey that can be used as economical upgrades?

I would like to get everything from 1 or 2 suppliers if possible, and it looks like I will be ordering a few CL90's, uF4007 rectifiers, and a filtered IEC socket from digikey as I do not see these items on the AES website.

Thank you,
Spraiski

[mesherm]

here is the schematic for a K-12 that uses 11MS8s and is almost the same.
http://webpages.charter.net/mesherm/k12m_schematic.jpg
I would suggest building it with the stock parts first and listening to it for a while. I doubt that just changing the coupling caps will make much of an audible change. Mine came with Xicon polypropylene caps which are decent sounding.
The most effective mods are to change the output transformers to Hammond 1608s and converting to UL, adding capacitance to the power supply and either rectify or reroute the filament circuit traces to reduce hum.

If you are determined to change the caps then either use
C-RD22-600 (715P orange drops)
C-PD22-400 (716P orange drops)
C-FSD22-630 (Solen fast caps)
all available from AES
Assuming the circuit is the same you will need 6 of them
P-QFR107 is the AES PN for fast recovery diodes

here is the definative K-12/k-502 mod web page
http://www.siteswithstyle.com/VoltSecon ... _Pull.html

[EWBrown]

I found that the WIMA 0.15uF 400V MKP caps fit the PCB holes perfectly. Search around the site, there are LOTS of postings relating to these amps.
THe 11BM8 version is more amenable to Ultralinear operation, the 11MS8 / 10GV8 version isn't.

AES used to carry the WIMA caps, not sure if they still do or not...

Update:

The WIMA caps that I used were the MKP10 0.15 uF, 400VDC, 250VAC.
AES no longer has these listed. It looks like they are just letting their remaining WIMA stock dwindle off into obscurity. Too bad, these are nice caps, and very reasonable price.

My next K502 experiment will be to try a 190VAC. 2oo mA & 26VAC @ 700 mA, toriodal power trannie from Trafomatic to power one of these. THe 26VAC, rectified will provide around 32VDC< and I'll re-configure the filament connections to run 16A8s in series.parallel.

/ed B in NH

[Spraiski]

Thank you for the capacitor suggestions.

I'm sure that I'll be doing a bunch of other modifications in the future, but I've got to start somewhere The rerouted filament traces will probably be included from the start as well.

If I find that the caps don't make a big difference the worst case scenario is that I have a few extra caps in the junk bin...

Spraiski

[TerrySmith]

I have both versions, the K-502 and K-12M. Both sound extremely good considering the price. Both suffer from slight background hum, not too bad.

The boards are not very good, traces seem to lift easily, so if you want to mod it I would build it point to point on a metal chassis.

[EWBrown]

That "incurable" background hum is from the AC filament traces being located between the pins of the interstage coupling caps. They could have not chosen a worse place to locate them... The PC board layout isn't very good, and I have been in contact with Boris Sasic over at Audio Vista Engineering to design an inproved PC board design - he already has a SET / SEP board which can handle either two or four 16A8s, and I thought of taking that board and dooing some etch surgery, but that proved out to be more of a task than it would have been worth...

The only real cure is either DC filament voltage -the supplied power trannie filament winding won't like this, as there will be about 42% additiional power draw, so an external trannie will be needed. When I modified to using 16A8s, I added an extra trannie, and a DPDT slide switch, mounted under the chassis so either the original 11 VAC or 16VAC could be selected., allowing 11BM8s or 16A6s. I chose the re-routed filament leads rather than DC in this version.

FWIW, I notrice that the actual AC voltage on the original unmodofoed K502 tended to run a bit low. I did use a CL-90 in series with the primary, which drops about 3.6VAC off the primary voltage, but then my AC line voltage tends to run a bit high, between 123-128VAC depending on time of day, etc.

I chose using 16A8s, since they are available as NOS for low cost, I'd find a sleeve of five NOS Raytheons on E-bay for $10, and at local swapfests for even less. Once in a while the Westinghouse brand, made in Great Britain NOS 16A8s are available, for about 2X the cost, and those are quite nice! I dn't know if they are Mullard, Brimar or what other UK tube company may have produced them.

/ed B in NH

[Spraiski]

I have successfully built my first tube amp!

I used a hammond chassis to keep down noise and keep up looks.

While building it, I performed the following mods:
-CL90 in series with the PT primary
-switched the coupling caps for 716P orange drops
-added a single QFR-107 fast recovery diode after the B+ bridge rectifier
-increased B+ capacitance with the same 330uF Nichicon caps used on the VoltSeconds mod page
-1uF snubber caps on the filaments of all 4 tubes
-0.01uF cap across the HV transformer leads


Next will probably be a DC filament supply. I'm thinking of building a variable regulator so that I can try different tubes.

I am also thinking of switching the output transformers for an ultralinear hammond. The circuit diagram of the K-502 is essentially the same as this K-12 schematic posted earlier.


I'm not sure of how to connect an ultralinear transformer (such as the Hammond 1608/1609 shown below) in this design. I think that the Blu (P2), Red (Ct), and Brn (P1) wires are connected the same way as the original OPT, and that the Brn/Yel Sg wire goes to pin 7 of each tube, but I'm not sure of what other changes are needed.


Spraiski

[mesherm]

The polarity of the Hammonds is reversed from the stock trannies.
The brown leads go to the solder terminal closest to the PCB edge and the blue leads go to the inside pads. Red stays the same.
You are corrent about where the UL tap goes. The blue/yellow goes to pin 7 of the same tube as the blue wire is connected. Same with thye brown/yellow. It goes to pin 7 of the brown leaded tube. Be sure to cut the circuit traces that conduct voltage to all the pin 7s. You want each pin 7 isolated before you solder the UL wire. I have read that some people even put a 100 ohm resistor in series between the screen grids (pin 7) and the UL taps but mine is wired directly and works just fine.

[Shannon Parks]

Looks cool and nice mods.

Thoughts:
-I would consider putting a 300k to 470k 1W bleeder in the power suppy with all that capacitance. Some B+ voltage will stick around when off and might bite you.
-FYI, the wrong Hammond polarity can cause a crazy oscillation once you fire it up with your feedback applied. Luckily, you can just power down and switch the polarity when those things happen.

[EWBrown]

The 11BM8 / 16A8 K-502 and the 11MS8 / 10GV8 K-12 (or 12-M or S-5) have very similar circuitry, just some different resistor values, and the different tubes have completely different pinouts. Same "UTK" transformers are supplied with both kits.
UTKs are made Italy, I wonder if they had "UTC" in mind...


George Fathauer is the "father" of these little amps, and yes, they go by many names.

The coupling caps can be 0.1 uF or 0.15 uF / 400VDC. 250VDC is cutting it close, and 600V is an unnecessary luxury

Bypass the two 68K NFB caps with 33 or 39 pF SM caps, or Polystyrenes.

Bypass the output stages' 300 (or 200) ohm resistors with 220-470 uF / 35V electrolytics.


/ed B in NH

[Spraiski]

It seems like a good time to resurect this old thread, so here goes

Since my last post I have bypassed R8/R17 with 470uF electrolytics, and added a 470k 1W resistor to the power supply. I still haven't gotten around to the upgraded OPTs as I've been too busy enjoying listening :D

I'm planning on building something like the diyparadise simple EL84 sometime in the near future (yes, I'm hooked), and I think a pair of Edcor or Hammond OPTs for the K-502 will probably come in along with an order for parts for the next amp...which actually brings me to my next query.

Can the CCS circuit suggested for the Simple EL84 (http://www.diyparadise.com/yhlmccs.html) be used in the K-502? Would it simply be a matter of taking out the 300ohm resistor and replacing it with the CCS circuit?

[DerekVa]

This is a timely resurrection...

I'm in the process of (re) doing a K12 / K502 as well for my office (currently I'm running a K501 w/ a single FE103 driver and want to move into the stereo age). I'm planning on doing the majority of the Voltsecond mods (sans CCS - I just don't understand how those things work), and am debating replacing the OPTs. My big question is how do the Edcors compare to the Hammonds? Are they the best bang-for-the-buck?

-D

p.s. I'm still (also) working on my P2P ST35 (rebuild w/ exotic parts and better P2P layout - original layout looks rat-nestish), RCA 12AX7 phono preamp and FE103 reflex cabinets so this may take a while - I'm not as productive / efficient as Tom! Yellow_Light_Colorz_PDT_07

[Gingertube]

Spraiski,
The addition of the 470uF Cathode Bias Resistors Bypass Capacitors was a good thing to do. It reduces distortion considerably when the output stage changes from Class A (both tubes conducting current) to Class B (when one tube will cut off - 0 current - on signal peaks).

The "Simple EL84" is quite elegant in that it uses the output stage itself as the phase splitter. It does this by putting a current source in the cathode circuit instead of the cathode bias resistor. This fixes the current through the 2 output tubes, when ever one conducts a little more then the other must conduct a little less. Unfortunately this method forces strict Class A operation (You can drive it hard enough that one tube will cut off BUT the other tube can never conduct more than what the current source is set for, which is 2 x the idle current of one tube) and so power output will be reduced by a fair amount.

Other schemes exist and over the last couple of years I have given most of them a try while developing and refining the "Baby Huey" amplifier.
For reference the whole "Baby Huey" saga is here:
http://www.diyaudio.com/forums/showthre ... adid=72536

The latest version can be seen on post #431 on page 18 of that thread.

I started with separate current sources in each push pull cathode, and each bypassed with a capacitor similar value to what you are using. This gives the benefit of enforcing balanced idle currents which will always help the bass end - while still allowing Class AB operation.

The lastest version, what I'm running at home now, is similar BUT I have reverted to fixed bias for the output tubes.

In the current source biased version I tried every bypass capacitor type going, and was reasonably satisfied which 1000uF/25V Blackgates BUT in the end found that best results came from fixed bias and thus eleiminating the need for any bypass capacitor.

Hope there is something of value to you in this rave.
Cheers,
Ian

[EWBrown]

The K-502 and K-12 PC boards have some strange grounding issues, and they both run the AC filament voltage underneath the interstage coupling caps, so there can be a seemingly incurable low level background hum problem. From my experience with this, in several amp kits that I have built, each individual case seem to have its own individual cure for reducing or eliminating this hum.

If the output stages are purely Class A, then the shared cathode resistor bypass cap isn't needed, and would have little effect, but as most of these circuits are really Class AB1 or AB2, then the cap really helps, in improving the bass response and lowering the 2HD and 3HD content.

Class B, as indicated by Gingertube, has one side of the PP circuit cut off, while the other side is conducting, and this will introduce the nasty little audio gremlin known as "crossover distortion", so Class B isn't used for many audio circuits, other than high-powered AM transmitter plate modulation, in which case, true high-fidelity sound quality isn't the major concern.

Class B is more power-efficient, as the idle current is low when there is no signal being amplified, but at the cost of increased THD and IMD.

/ed B

[Spraiski]

Well I have managed to evade the hum with my K-502 (or it might just be my horribly inefficient speakers...). I did add 1uF snubbers on across the filaments on each of the tubes and suplemented the filament traces with a pair of tightly twisted wires connecting the filaments on the two outer tubes.

btw, is there an easy way to determine if an amplifier is purely Class A wwhen looking at a schematic?