Looking for varistor V023L4

[romainbrette]

Hi everyone,

I'm looking for a varistor VECO 023L4, manufactured by Victor Engineering Corporation in the 60s/70s. Unfortunately, the company doesn't exist anymore and I couldn't find the specifications of the component anywhere. I need this to build an analog circuit modeling a neuron (from an old paper written in the 60s), and it's crucial that the component has similar characteristics. Would you have any idea where I might find the component, or a similar one, or any information about it?

Thanks a lot!
Romain

 

[BradtheRad]

Don't know if varistors were ever used for much else than absorbing spikes and surges in power supplies.

To build a circuit using a varistor implies you're running at voltages over 130V. That's the typical point where a varistor activates (these days).

Can you consider using a lower supply voltage? Then you could use zener diodes. They are inactive below a certain voltage, like varistors. Above that voltage they conduct, like varistors. Two zeners back to back should behave very much like a varistor, within limits.

Moreover, varistors aren't meant to turn on frequently. They start going bad after absorbing a number of surges. They may be ruined by a single powerful surge.

So I wonder if a varistor is really what the author meant should be used in a circuit? Was he speaking theoretically?

Did he use tubes in his circuit? Tubes use high voltages.

Can you scale down the voltage levels for solid state components?

==========

After checking I see there are low voltage varistors available mail order.

And it's true that varistors can handle higher current than zener diodes. Yet there are high watt zener diodes.

Don't know how expensive it would become if you were to use one or the other.

====

Another component to consider is the diac. Often used in light dimmers to trigger a triac. Triggers around 30v, then will conduct at lower V for as long as V remains above a certain level. I see these are available at a few mail order houses.

Or a neon bulb. Triggers around 60V. Low current only. These are sometimes used instead of diacs.

 

[romainbrette]

Hi,

Thanks a lot for your message. Actually what I'm trying to implement is this: https://www.eecs.berkeley.edu/~lewis/ProcIEEE68.pdf
specifically the circuit p943. I could find all components except the N1, which is that varistor. There's a 12V voltage supply in the circuit. But the varistor is really used as part of the neuron model. Specifically, it models the inactivation of the sodium channel. This is a voltage-dependent current, and the inactivation can be seen as a conductance that decreases when the voltage increases. It is required to make the neuron potential go back to its resting value after an action potential is initiated. There's another paper where he shows what I interpret as the voltage-dependence of the conductance of the varistor (given my very limited knowledge of electronics), I'll try to post a link.

Romain

 

[BradtheRad]

Looking at the circuit schematics, I notice practically all components have a battery icon in series.

I think that the author found it necessary to add some inline voltage to each varistor in order to bring the varistor close to its trigger point. Probably needed for the sake of increasing sensitivity. The circuit might not function except for a slew of batteries.

I also think the author found it necessary to custom tweak the voltage inline with each varistor. To precisely the right voltage. No doubt you'll have to do the same.

In other words, component values can vary from the parts list. I think you can use varistors available today. Because even if you could obtain the V023L4 varistors, you'll have to do a lot of work to make the circuit operate. Notice the parts list calls for potentiometers. They will need you to adjust them.

I see why the author did not use zener diodes. They turn on more gradually, rather than trigger.

I don't know how close the battery voltage should be to the varistor trigger point. The only easy way to vary battery voltage is by the number of cells. You may have to mix and match different quantities of 1.5V alkalines, 1.2V nicads, etc.

You might consider obtaining varistors with a variety of trigger voltages. (Example, 5.5, 7, 9V are available from mouser.com, a reputable electronics mail order supply house).

...Sorry that was Digi-Key, but also check Mouser:

www.mouser.com/Circuit-Protection/Varistors/_/N-5g3h/

---------------------

Looking at the schematic on p. 943.

I see that rather than each varistor having a battery inline, there's a nearby potentiometer to provide tailor made adjustments.

 

[romainbrette]

Hi,

Here's a paper where I think they give the current-voltage relationship for the varistor: http://audition.ens.fr/brette/tmp/Lewis68.pdf
Fig 9 on page 37 shows the conductance vs. voltage, but I believe the voltage has to be scaled up by about 100. They explain the circuit a little in section F ("sodium conductance"). Apparently they can tweak this curve with a potentiometer.
Do you know a modern component that would behave similarly?

Romain

 

[BradtheRad]

Fig 9 on page 37 shows the conductance vs. voltage, but I believe the voltage has to be scaled up by about 100.

The graph appears to show the trigger action at -15mV.
(Not sure whether the negative sign is absolute, or in reference to a more positive value.)

Scaled up 100x results in -1.5V as the trigger (or 1.5V? The spec for varistor trigger V is a positive value). This may or may not be the spec voltage for your varistors.

Anyway the action takes place over a span of 50mV. Scaled up 100x gives you 5V as the span of action. This may or may not be the varistor spec you need. I imagine it should be less than the supply V which is 12V.

The Mouser webpage shows a large selection of voltages. Not every item is necessarily in stock. It may be wise for you to purchase an assortment.

No one can recommend which type will work in your circuit, since this is not a typical use of varistors, nor is it a typical circuit.

Can you find sufficient instructions from the author for duplicating their experiments? They may have left out a few crucial details. Such as the trigger voltage for their varistors. And how to adjust the potentiometers.

Did they have their reasons? That's where human nature enters in.

 

[romainbrette]

Thanks. The author gives some indications about setting the potentiometers, and also I have a rough idea of how the currents should look like if they are supposed to match physiology, so I guess I could play a little with the potentiometers.
For the varistor, thanks for the link, the only thing is I don't know much about electronics so I'm not sure how to select the different values... I can see that there are three different voltages, could you explain me what they mean? Also, are the "peak surge current", "surge energy rating" and "capacitance" important? (and is there an important difference between MLV and MOV?)
Sorry about all these questions, I'm really a newbie...

 

[BradtheRad]

The chief spec you're interested in is maximum continuous DC working voltage.
Probably 3 to 9V.
I'm assuming it should not be greater than the supply voltage shown in the schematics.

The other parameters are unlikely to be a major consideration. I'm also assuming AC operation is not pertinent.

I looked at a few mail order websites. Digi-key, Mouser, Jameco, Hosfelt, All-electronics. Many have quirky selection tables. Hard to narrow down to the specs you want. Probably have more success calling them on the phone.

Low voltage MOV's are commonly surface mount devices. Not easy to work with. Tricky to solder.

One MOV that turns up is the V8ZA2. The clamping voltage of 22V may or may not make a big difference for your purpose.

As for MLV, it's Multilayer Varistor. Specifically designed for suppressing electrostatic discharge.

Has someone done an updated version of the experiment in the past 40 years since the article came out? Perhaps there are components better suited for your purpose.

Possible prospects:

* avalanche diode, trigger diode, etc. (These are not to be confused with ordinary diodes.)

* transient surge suppressors (transorb)

* silicon controlled rectifier (scr). Conducts when a high enough voltage is applied to gate, and conducts until current drops to zero.

* unijunction transistor (ujt). Usage has declined over the years so it's hard to find in one inexpensive package. Can be duplicated using two transistors.

* Programmable Unijunction Transistor. Two transistors arranged to act like a UJT.

These components may or may not exhibit the precise action to make the experiment work. Someone more expert than I would give you better advice as to what's feasible to work with nowadays.

Can you determine what kind of performance graph is needed?
Particularly...

Whether the component should activate and shut off above a set volt level (as with varistors)?

Or whether, once in turns on, it should switch off at a lower volt level than it turned on (as with a diac, UJT)?

Or whether, once in turns on, it should continue conducting until current falls to zero (as with an scr)?

 

[romainbrette]

Thanks a lot for all this information!
Nowadays, electronic neurons are built quite differently, but unfortunately people tend not to publish all the details... Actually I just contacted the author, but unfortunately he told me he could not find any information about the varistor. I'll try with your advice!

 

[BradtheRad]

Best of luck.

Researchers seem inclined to give us just enough details showing how they got their results. However they won't be anxious to 'give away the store', as it were.