Varistor Characteristics for Magnatone Vibrato Circuits

[Smokebreak]

Hey Tim, it's great to see you in the discussion. I've been doing the audio/recording tests, and exhanging subs with Martin for a while now.
I've been using a 213 clone.
I'd be more than happy to give yours a run in this circuit if you are interested.

Cheers- Jeremy

[martin manning]

Jeremy, above I proposed replacing the Zeners with modern MOV's as a possible cure for the clicking that made the Zener version unacceptable. If you could try that it would be great. I don't have any of the Littlefuse MOV's, unfortunately, so you'd have to get some. Or, you had a bunch of one voltage rating as I recall (60-ish volts?), so maybe you could make up a couple of units with just one break point to see if the clicking appears?

[trobbins]

Jeremy - did you only try Martin's original clone circuit, or the later version with a 33V zener parallel arm? I'd certainly suggest you try and test that variant in your amp for starters, as it is the variant I did measurements on.

It may take a day or two for me to test the 80-100V MOV I have. One practical hassle with MOVs may well be their tolerance at around 1mA, which I would think would be far worse than zeners, and hence tuning the varistor clone with resistors may not be 'universal'.

Ciao, Tim

[Smokebreak]

I've tested 4 variants so far :

Workman NOS : delightful

Littlefuse MOV : worked, but produced tremolo

Martin's original "earring" zener sub : https://tubeamparchive.com/files/varist ... te_141.pdf : produced vibrato but with clicks

NOS Russian SIC 100V : worked well, a little thumpy. Martin has matched these and I will be re-testing soon, as the ones out of my lot I randomly selected were apparently the worst matched.

I haven't tried either of Martin's latest, be it the MOV sub or the long zener strings(Tim, your original or updated). It would be helpful for me to see schematics for basically everything I haven't tested, as I'm unsure where you guys are shunting, and I'm starting to get a little lost in varistoresistor land !

[martin manning]

My put would be to try the MOV variant posted above as I think it's the best bet. Adding more Zeners (i.e. a break point at 33v) might more closely approximate the curve, but if it still clicks or otherwise muddys the sound then it's not satisfactory IMO. The long Zener string is not my idea, but there the part count is very large. I don't recall any mention of clicking with the MOV in the previous test, just that it was more like tremolo.

[Smokebreak]

My put would be to try the MOV variant posted above as I think it's the best bet. Adding more Zeners (i.e. a break point at 33v) might more closely approximate the curve, but if it still clicks or otherwise muddys the sound then it's not satisfactory IMO.

Will do. It may take me some time to get to this.

I don't recall any mention of clicking with the MOV in the previous test, just that it was more like tremolo.

You mean the little red ones? Yep, no clicking, just tremolo. My idea there was to have a vib/trem switch for the user to decide

edit : Martin, Can you elaborate on the "break point at 33V? I'm not following here.

[martin manning]

Tim saw a strange s-curve in the I-V response he measured, and got rid of it by adding another Zener with 33V breakdown voltage. I don't understand how that came about, and I didn't see it when I traced the original version which I sent to you. In any event each voltage and resistor combination adds another straight line segment and slope, which means you can more closely match a SiC varistor curve. With three breakdown voltages and four resistors you get four slopes.

[martin manning]

OK Jeremy, let's try this MOV thing... I ordered parts from Mouser, and I'll make up a couple and send them down. Note I read the MOV data sheet a little more closely and adjusted the MOV part numbers and resistor values to get a better match to a SiC varistor curve (see above).

[Smokebreak]

Sounds like a plan!

[trobbins]

I tested a MOV last night, and made up a schematic of the clones I tested, but haven't updated doc yet. Over the current range 0.1 to 1mA, I recall that the MOV voltage increased from about 90 to 100V. In comparison an 82V zener voltage is static across the 0.01 to 1mA span. So if MOVs can be got with suitable voltages at the current levels of influence for that MOV then the degree of wobble along the V-I curve should be softer.

[trobbins]

Attached is updated measurement result doc with schematics of the varistor clones used for the results. The 33-56-82V clone is I reckon worth trying in an amp due to its simplicity.

I prepared the 3V6 string clone because 5V6 and above zeners have a very static junction voltage for currents between 0.01mA and 1mA, whereas lower voltage zeners (well 3.3 to 3.9) have quite a variation in voltage, which somewhat approaches the SiC varistor variation.

I appreciate the 3V6 zener clone is tedious to make (given that I have prepared two pairs of them), but had a view that if it did result in a good approximation then someone would get keen and ask a Chinese job shop to make a batch of varistor pairs using smt parts at $2 a pop.

[dinkotom]

Guys, of course I can't understand a word, but I'm keeping fingers crossed and appreciating the hard work your're doing for paint-by-number types like me, so we can have affordable Maggie vibes. Every house should have one!

The originators:
http://www.youtube.com/watch?v=nqHoAkZDG5w
http://www.youtube.com/watch?v=bTIucAz_jbg

[martin manning]

Nice summary, Tim!

There are perfectly good SiC varistors being produced, they're just not available in small quantities at a reasonable price. That is also the case with NOS parts. If this 3x MOV replacement works out then then there will be an alternative, with parts cost currently at $2/unit.

[martin manning]

Faux SiC Varistor part deux: MOV's replacing Zeners.

Breadboarded two MOV-Resistor units to see how well they match design-intent and each other, with very good results. This looks promising! The measured data from the curve traces looks jittery at low current due to rounding, but shows very good agreement between the two units. Next: solder them up and do a listening test.

[andyfromdenver]

Here's some eye candy for the brains around here
200 gallons of effort later, and I'm about to finally build this lil guy.
If you've been there, you know how satisfying it was to finally push in the button lamp

Thanks so much all for your varistor effort and sharing!

I've got a build thread going elsewhere on the web, with transformer details, musings, etc. cause I didn't know the etiquette of where non FMV&T&D go here. I'll come back with a demo video once it's warbling.