Introducing the 70s USSR era Kazan "Vibrato"

[Paul Marossy]

This is my webpage on it: http://www.diyguitarist.net/DIYStompboxes/Kazan%20Vibrato.htm

I think it is a super cool and good sounding tremolo in a package that makes it very useful in phrasing. I am not really one to use tremolo, but this one suits my style well since it's like a gas pedal - press down the treadle and you get more, let off entirely and you get none. Goes from very subtle to very choppy. I had to wait two months for it to arrive from Kazakhstan but it was worth the wait. Was NOS in the box. I had to convert it to 1/4" jacks from the original 5-pin DIN jacks and while I was at it I converted the wiring to switch power at the input jack instead of the output jack.

The circuit design is very interesting, uses FETs and germanium transistors - NPN and PNP. Scope shows a little bit of asymmetrical clipping with higher input levels.

[Rob Strand]

FYI, on the schematic,  the base of T4 should also connect to the node where R12 and R13 join.

The way the schematic is drawn doesn't help understand how it works - yeah I realize the manual draws it like this.

This makes more sense signal flow wise,

T3 -->  divider R11, R10 ---> T6 (in base, out emitter) --->  JFET ---> T4 (in emitter, out collector)

T4 operates as a common base amp.

The JFET resistance forms a divider with the low input impedance at T4's emitter.  That's the primary tremolo mechanism.

The bias voltage at the emitter of T6 is seems critical to biasing the JFET, especially when the Depth is set to minimum.     With the depth is advanced you might get some re-biassing if C6 charges through the JFET's gate diode; like the EA tremolo.

Hard to believe NOS from Russia and that era - amazing!

[diffeq]

KT315 is notorious for having fragile leads that snap at the body when mishandled.  It is silicon npn.

The manual might have slight errors - schematic shows bipolar supplies but text mentions only one battery, innards also have only one 9v snap.

[Rob Strand]

Early on today I dug around for info on the KB303/KP303.   It appears the KP303 is sorted into groups like the 2SK30A etc..   From the manual it looks like the circuit specs a кп303ж.  ж translates to "ZH" and from what I can see it doesn't have a simple western suffix letter.  Also it's not that common.

I think the KP303ж parameters worked out around to be the low VP and IDSS end of the range; roughly VP=1V and IDSS=0.8mA; spice beta = 0.8m.

I suppose common parts like the 2N5484 or J201 are in the same zone of parameters.

From the common suffix parts the KP303ж was closer to the KB303A.

Anyway I dug-up these spice models which seemed to have similar parameters to what I expected.

.MODEL KP303A-RUS NJF VTO=-0.9 BETA=1.6m LAMBDA=20m RD=30 RS=35 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR
.MODEL KP303B-RUS NJF VTO=-2 BETA=2m LAMBDA=20m RD=30 RS=35 CGS=5p CGD=4p PB=1 IS=100f AF=1 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=1f M=0.5 B=1.2 NLEV=2 MFG=USSR
.MODEL KP303G-RUS NJF VTO=-2.5 BETA=1m LAMBDA=25m RD=22 RS=17 CGS=5p CGD=4p PB=0.5 IS=115f AF=1.3 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.07f M=0.5 B=1.2 NLEV=2 MFG=USSR

When I did a quick look at the circuit the bias voltage on the emitter of T6 seems higher than expected based on those JFET VP values.   Haven't had a chance to do any more checking.

[Paul Marossy]

FYI, on the schematic,  the base of T4 should also connect to the node where R12 and R13 join.

The way the schematic is drawn doesn't help understand how it works - yeah I realize the manual draws it like this.

Yeah, that improves the waveforms in LTSpice. I couldn't tell from looking at the schematic they're supposed to be connected and I don't know enough to know any better.

This makes more sense signal flow wise,

T3 -->  divider R11, R10 ---> T6 (in base, out emitter) --->  JFET ---> T4 (in emitter, out collector)

T4 operates as a common base amp.

The JFET resistance forms a divider with the low input impedance at T4's emitter.  That's the primary tremolo mechanism.

The bias voltage at the emitter of T6 is seems critical to biasing the JFET, especially when the Depth is set to minimum.     With the depth is advanced you might get some re-biassing if C6 charges through the JFET's gate diode; like the EA tremolo.

So is the FET acting like a switch that ramps on/off at regular intervals or is it C8 charging/discharging? Or is the FET like a variable resistor? It's a very quiet circuit, no hiss even with headphones.

Hard to believe NOS from Russia and that era - amazing!

Yeah, my Kvaker wah by same manufacturer (another interesting circuit I shared here a few months ago) was also NOS but missing the the box top. That one came from Ukraine. Amazingly it got here in spite of the war going on. Usually if I dig long enough I can find some stuff about obscure manufacturers but in this case I can find absolutely nothing. Not surprising I guess considering where it comes from. I sure would like to know who designed it, and how manufacturing that worked in that era... like was it state sponsored thing? And so on. The curiosity is killing me.

Thanks for the models on the KP303. I didn't have one so I used something that is supposed to be somewhat close to it.
EDIT: I just tried it with the KP303A model and that looks remarkably close to what I saw in real world on the scope!

[Paul Marossy]

KT315 is notorious for having fragile leads that snap at the body when mishandled.  It is silicon npn.

The manual might have slight errors - schematic shows bipolar supplies but text mentions only one battery, innards also have only one 9v snap.

It's not a dual supply, it's just a 9V "positive ground" circuit. The jack wiring is a bit confusing with those 5-pin DIN jacks.

[duck_arse]

KP303B (rus: КП303Б) and KP303A (rus: КП303A) can be had from Ivan at diy-tubes.com.

[Rob Strand]

So is the FET acting like a switch that ramps on/off at regular intervals or is it C8 charging/discharging? Or is the FET like a variable resistor? It's a very quiet circuit, no hiss even with headphones.

The voltage on C8 remains fixed.  C8 and the two voltage divider resistors act like a Vref circuit on an opamp pedal except it biases the bases of the two transistors T4 and T6.

The JFET acts like a variable resistor.  When the sine output from depth control is more positive the JFET is on and the signal passes.  When the depth is off the gate is actually forward biased and the JFET is fully on.   So the small detail is when the sine wave from the depth control swings positive it also forward biases the gate and passes the audio signal.  What happens through is once the oscillator has been running for a short time the cap C6 charges up through the gate diode of the JFET on the positive peak of the sine-wave.   At the positive peak the JFET is fully on (just).  That means the negative swing of the sine wave the JFET is driven  off.   That's how you get the modulation.

The EA tremolo works in a similar way in that there is a cap between the JFET gate and the oscillator output.  In the case of the EA it's the cap between the oscillator and the depth pot.

The charging cap means the biasing on the emitter of T6 *isn't* that critical.  I was wrong before thinking the emitter was important as I was looking at the circuit as if the emitter voltage was positive like an NPN circuit, but it's actually negative which drives the JFET on when the depth is on zero.

The advantage of the C6 charging on the oscillator peaks is the only thing that changes with different JFETs is how far you have to crank the Depth control to get the signal to pinch off (the dips in the modulation).   A low VP JFET requires a lower Depth setting than a high VP JFET to get the same amount of pinch-off.   From what I can see with a few different JFETs is the depth control for pinch-off is somewhere between half and full.

Yeah, my Kvaker wah by same manufacturer (another interesting circuit I shared here a few months ago) was also NOS but missing the the box top. That one came from Ukraine. Amazingly it got here in spite of the war going on. Usually if I dig long enough I can find some stuff about obscure manufacturers but in this case I can find absolutely nothing. Not surprising I guess considering where it comes from. I sure would like to know who designed it, and how manufacturing that worked in that era... like was it state sponsored thing? And so on. The curiosity is killing me.

I noticed that the other day.  I had a quick look at the schematic but had to do something else.   Looks cools,  another inductorless wha.


Something else I forgot to mention.  If you look at T3 it looks like a buffer *but* because of the positive ground it's actually a gain stage.   Technically the drain and source are drawn reversed but it still works because JFETs work with d and s swapped.

[Paul Marossy]

Yeah, my Kvaker wah by same manufacturer (another interesting circuit I shared here a few months ago) was also NOS but missing the the box top. That one came from Ukraine. Amazingly it got here in spite of the war going on. Usually if I dig long enough I can find some stuff about obscure manufacturers but in this case I can find absolutely nothing. Not surprising I guess considering where it comes from. I sure would like to know who designed it, and how manufacturing that worked in that era... like was it state sponsored thing? And so on. The curiosity is killing me.

I noticed that the other day.  I had a quick look at the schematic but had to do something else.   Looks cools,  another inductorless wha.

Honestly it's one of my favorite wah pedals now. Has a very distinctive sound with an overdrive/distortion pedal. I like how, circuit-wise, it's also completely unlike anything else out there.
I don't want to play it too much because a replacement pot is virtually impossible to find but hard not to. 

[Rob Strand]

Honestly it's one of my favorite wah pedals now. Has a very distinctive sound with an overdrive/distortion pedal. I like how, circuit-wise, it's also completely unlike anything else out there.
I don't want to play it too much because a replacement pot is virtually impossible to find but hard not to.

I always find it interesting how some of the circuits from Russia (and surrounding areas) don't follow the usual circuit pattern.

I tried to cobble together some spice models but I only got this far:

KP303


Local library for this pedal.  KV.LIB

Code Select Expand


* Some approximate models for parts used in the
* Russian Kazan Vibrato
* Rev 1.0, 22 July 2023

* Generic PNP germanium model
* hFE ~ 65, leakage Iceo ~ 100uA
.model MP26B_KV pnp(Is=9.40u Bf=55 Nf=1.0 Vaf=55 BR=7 Var=50
+                 RB=25 RBM=0 RC=1 EG=0.68  XTI=3.0
+                 ISE=0 ISC=0
+                 VJC=0.7 MJC=0.333 VJE=0.7 MJE=0.333
+                 Cjc=100p Cje=200p TF=106n TR=140n mfg=USSR)

* The NXP 2N3904 model with reduced BF
* hFE ~ 200
.model KT315B_KV npn(IS=1E-14 VAF=100 Bf=184
+ IKF=0.4 XTB=1.5 BR=4 CJC=4E-12 CJE=8E-12
+ RB=20 RC=0.1 RE=0.1 TR=250E-9 TF=350E-12
+ ITF=1 VTF=2 XTF=3 Vceo=40 Icrating=200m mfg=USSR)

* LTSpice model for A suffix
.MODEL KP303A_KV NJF(VTO=-0.9 BETA=1.6m LAMBDA=20m RD=30 RS=35 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR)

* Approx model for "Z" suffix used in the Vibrato
* VP= 1.0V, IDSS=1.4mA, rds_on=360 ohm
* (ID = 1.62mA @ VDS=10V VGS=0)
.MODEL KP303Z_KV NJF(VTO=-1.0 BETA=1.4m LAMBDA=20m RD=30 RS=35 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR)

* Alternative *artificial* JFET models for with higher VTOs for experimentation
*
* VP= 1.5V, IDSS=2.78mA, rds_on=270 ohm => beta = 1.23m
.MODEL KP303Z1_KV NJF(VTO=-1.5 BETA=1.2m LAMBDA=20m RD=27 RS=32 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR)
*
* VP= 2.0V, IDSS=4.55mA, rds_on=220 ohm => beta = 1.14m
.MODEL KP303Z2_KV NJF(VTO=-2.0 BETA=1.14m LAMBDA=20m RD=24 RS=28 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR)
*
* VP= 2.5V, IDSS=6.58mA, rds_on=190 ohm => beta = 1.05m
.MODEL KP303Z3_KV NJF(VTO=-2.5 BETA=1.05m LAMBDA=20m RD=22 RS=25 CGS=5p CGD=4p PB=1 IS=100f AF=1.4 FC=0.5 BETATCE=-0.5 N=1 NR=2 XTI=3 VTOTC=-2.5m ALPHA=1u VK=1 ISR=10f KF=0.1f M=0.5 B=1.2 NLEV=2 MFG=USSR)



From what I can see from the Russian manual the pedal specs (english) B variants the MT26B and KT315B.

[Paul Marossy]

I rounded up a SPICE model for the MP26A and I had found one for the P416B when I modeled the Kvaker a few months ago. I found something that said the KT315B is basically same as the 2N2222, so I used that. Used the KP303A model provided in this post.

These might be helpful? I don't have one for the KT315B, yet. Google app works most of the time to translate them.

MP25A/MP26A, etc. datasheet: http://www.diyguitarist.net/PDF_Files/MP25A_datasheet.pdf
KP303 datasheet: http://www.diyguitarist.net/PDF_Files/KP303_datasheet.pdf
P416A datasheet: http://www.diyguitarist.net/PDF_Files/p416a_p416b.pdf

EDIT: Just found the KT315 - http://www.diyguitarist.net/PDF_Files/KT315_datasheet.pdf  &  http://www.diyguitarist.net/PDF_Files/KT315A_datasheet.pdf

[Rob Strand]

I rounded up a SPICE model for the MP26A and I had found one for the P416B when I modeled the Kvaker a few months ago. I found something that said the KT315B is basically same as the 2N2222, so I used that. Used the KP303A model provided in this post.

These might be helpful? I don't have one for the KT315B, yet. Google app works most of the time to translate them.

MP25A/MP26A, etc. datasheet: http://www.diyguitarist.net/PDF_Files/MP25A_datasheet.pdf
KP303 datasheet: http://www.diyguitarist.net/PDF_Files/KP303_datasheet.pdf
P416A datasheet: http://www.diyguitarist.net/PDF_Files/p416a_p416b.pdf

EDIT: Just found the KT315 - http://www.diyguitarist.net/PDF_Files/KT315_datasheet.pdf  &  http://www.diyguitarist.net/PDF_Files/KT315A_datasheet.pdf

Yes, those are the best datasheets.  I got most of those except the last KT315A  datasheet.

Your MP25A/MP26A datasheet covers the (English) B version also.

For the MP25/MP26 and KT315 and manual specified the Russian Б suffix which translates to a B suffix in English.

The German KP303 datasheet I referred to in my summary comes from,
(Take note of how the Russian suffixes and German suffixes map in my summary.)
https://www.tiffe.de/robotron/Bausteinuebersicht/KP303.pdf

You probably have this MP25A model

Code Select Expand


* lt spice 2012
* hFE = 32, ICEO = 39uA
.model mp25a_lt pnp( bf=140 br=3 eg=0.72 cje=30p cjc=90p tf=0.4u tr=4u
+ is=3u ikf=40m vaf=30 mje=0.5 mjc=0.5 rb=120 rbm=40 irb=0.1m
+ ise=.3u ne=1.28 isc=.6u nc=1.28 mfg=GERMANIUM_USSR)


The gain of the A is quite a bit lower than the B.   Also leakage is pretty low but probably fine for a low hFE device, since the leakage is ICEO.

The model I gave has hFE = 65 which is more like the B suffix.   I didn't like that MP25A model because the observed gain is only 32 but the BF parameter is 140.  Some of the other parameters (ISE, ISC) are bringing the gain by a large amount-  too much for my liking.   The AC spice parameters look quite different between the two models but the capacitances end-up being fairly close in practice because of the TT parameter differences - close enough.

FYI: my previous list of models had MT26 not MP26, I fixed it.


Here's the result of a bit more effort at matching the datasheet
for the MP26B

Code Select Expand


* Target params (approx)
* gain: hFE ~ 55 @ 2.5mA & 20V, hFE ~ 65 @1.5mA & 35V
* leakage: Iceo=39uA @ 40V, Iceo = 50uA @ 70V
* capacitance: Ccb=32pF @ Vcb=20V, Ccb=30pF @ Vcb=35V
* fT = 750kHz @ Vce=20V, IC=2.5mA and fT = 750kHz @ VCE=35V & IC=1.5mA
* TR not tuned
.model MP26B_KV pnp(Is=3.25u Bf=40 Nf=1.0 Vaf=55 BR=7 Var=50
+                 RB=25 RBM=0 RC=1 EG=0.68  XTI=3.0
+                 ISE=0 ISC=0
+                 VJC=0.7 MJC=0.25 VJE=0.7 MJE=0.25
+                 Cjc=78p Cje=156p TF=210n TR=210n mfg=USSR)


[Paul Marossy]

This is the model I have for the MP25/MP26.

Code Select Expand

.model mp25a pnp( bf=140 br=3 eg=0.72 cje=30p cjc=90p tf=0.4u tr=4u is=3u ikf=40m vaf=30 mje=0.5 mjc=0.5 rb=120 rbm=40 irb=0.1m ise=.3u ne=1.28 isc=.6u nc=1.28 mfg=GERMANIUM_USSR)
.model mp26a ako: mp25a pnp (mfg=GERMANIUM_USSR)

It was in a bunch of files I downloaded at a SPICE related forum from someone named Bordodynov. I think that is also where I got the P416B too.

[Rob Strand]

This is the model I have for the MP25/MP26.

Code Select Expand

.model mp25a pnp( bf=140 br=3 eg=0.72 cje=30p cjc=90p tf=0.4u tr=4u is=3u ikf=40m vaf=30 mje=0.5 mjc=0.5 rb=120 rbm=40 irb=0.1m ise=.3u ne=1.28 isc=.6u nc=1.28 mfg=GERMANIUM_USSR)
.model mp26a ako: mp25a pnp (mfg=GERMANIUM_USSR)

It was in a bunch of files I downloaded at a SPICE related forum from someone named Bordodynov. I think that is also where I got the P416B too.

Yes, it's the same as the MP25A I posted earlier.   There's a standard.bjt file that comes from Bordodynov which contains those models.   

Bordodynov posted a large collection of models on the web which had all part types.  The standard.bjt file originates from LTSpice (model content varies in each LTSpice version with the earlier versions containing more models) then Bordodynov added models to that file.   There's many standard.bjt files on the web - hard to know if the file is the original LTSpice file or Bordodynov's.