5.333 volt

[JanJansen]

What is the best and cleanest way to get 5.333 volt with multiturn trimmer ?
I need to have exact 5.333 volt for building a TB303 synthesizer : http://privat.bahnhof.se/wb447909/dinsync/service_manuals/TB-303.pdf

[ajb]

Connect one side of the trimmer to >5.333V and the other side to <5.333V, and tweak it to suit?  Not sure what you're really asking for, here.  Maybe follow it with a unity-gain buffer if you need a lower impedance?  Or use a precision reference?

[alm]

If it is supposed to be a power rail, then an adjustable three terminal regulator like the LM317 would be the easiest solution. Feed it with something like 9 VDC.

But how about some proper requirements. Available supply voltages? Min and max current draw? Allowed noise? Budget? Volume?

[JanJansen]

It has 12v supply voltage, current draw i,m not sure.
I only make 1 or 2, what are the options ?

With the LM317 you have a resistor to ground, i dont want that to influence the schematic for the synth.

[alm]

Fine, then build the schematic from the synthesizer. It includes a power supply that creates 6 V, 5.333 V and 12 V rails from a 9 VDC input.

[technix]

You can match a 7805 regulator chip and a 1N5819 Schottky diode to give you the 5.333V. If properly matched the voltage drop of the Schottky diode will push the output voltage to 5.333V. Or from some less than reliabel brand (loose tolerances) you may even find 7805 chips with 5.333V output voltage.

[JanJansen]

Thats funny, buy a 5v regulator and hope for 5.333.
I,m not sure how to connect the Schottky diode.

[JanJansen]

I saw a : dont try this at home movie before,
someone hooked up a 7805, with a resistor to the ground, then it gives higher voltage.

[Zero999]

How accurate does it need to be?

The TL431BC is 0.5% accurate and 0.5% tolerance resistors aren't that expensive.


http://www.homemade-circuits.com/2012/12/explaining-programmable-shunt-regulator.html

For V_O = 5.333V

0.5% values
R1 = 1k82
R2 = 1k6

1% values
R1 = 1k87
R2 = 1k65

You could also replace either R1 or R2 with a slightly lower value, in series with a trimmer, but beware the drift of a trimmer will be higher than a precision resistor.

[JanJansen]

It has to be exact 5.333 volt measured.

[PA0PBZ]

It has to be exact 5.333 volt measured.

Exact doesn't exist and if it did you couldn't measure it.
So 10%, 1%, 0.1%?
And what is the accuracy of the DVM you are going to use to measure it?

[Zero999]

Yes, there's always a limit to precision. Nothing is completely accurate.

Another problem is drift. You might be able to trim to +/-0.001% which would be between 5.33294667V and 5.33305333V, it won't necessarily stay there. It may drift considerably over time, especially as the temperature changes.

[DBecker]

The original schematic (page 5 ref 25D) shows a 1sz46a zener creating a 6.2V reference, and then a resistor devider tweaked with a trimmer for creating a 5.333V level, with an opamp voltage follower creating the supply.

The service manual probably tells you to adjust to 5.333V at "operating temperature".  But it's pretty certain that it will drift significantly with changes in conditions and over time.  My guess is that you don't need the accuracy suggested.  They really mean "adjust to 5 1/3 V" and do a pretty good job, just like the 1V and 4V level are described as 1.000V and 4.000V

[sokoloff]

Between 5V and 6V; slightly closer to 5V; use your open eye. 

[Crumble]

I checked the schematic and it uses an 1SZ46A reference diode which is not accurate, but has a very low temperature drift. You will find the datasheet here. It is tweaked with a pot in the schematic, so if it says 5,333V, please tweak it with a 4 digit multimeter to 5,333, but do search for a 50ppm/C reference, because thats the spec on the 1SZ46A. It does not seem to be available, but you might use an LM329. It has a slightly higher temperature coefficient of 100ppm. The opamp (the AN6562) does not seem to be particularly temperature independent or low offset, but it does not have a major effect on the temperature drift of its output.

[JanJansen]

Thanks for the reply`s,
how many milliamps does it need for this schematic you think ?
Maybe i will just use 1 opamp and voltage divider, why zener ?
sorry i,m more beginner

btw : i still need to buy a multimeter that can measure this, for now i can set it to 5,33 volt.

Maybe i will a 6v regulator + a 5v regulator, then use the presision trimmer in between ?

[Crumble]

The zener requires only 1mA to work, and the LM329 works down to 0,6mA. I do not know how critical the voltage level on the syth is, it is quite difficult to tell how it works without the service manual. A synth will likely drift out of tune when something with poor stability would be used and I think the voltage will have to be tuned for a specific sound rather than voltage.

Creating the 5,333V from the 6V supply with a voltage divider would require an equally stable 6V supply, which this is not, because it seems to be fed from batteries that have a voltage that is dependent on quite a lot of circumstances, including their state of charge.

I still don't know what you're trying to do here. Do you have a synth of this type you are trying to repair, or are you trying to build a new one from scratch with the schematic? Given the specific part that was used, the designer seems to have gone for stability, else a cheaper part would have done. Please let us know what your intentions are, maybe with pictures and links to show what problems you are trying to solve.

[JanJansen]

Hi, i wanto create it from scratch.
I have many things todo, so i just sorting things out in advance.
I,m realy new to VCO stuff, no drift is realy important, i also did not solve on what to use instead of the 560 ohm tempco resistor.
Its hard to create a clone if parts are not available anymore, cloning the tr808 drummachine wil be easyer then this tb303.

[amyk]

What exactly does it need 5.333V for (and why wouldn't, e.g. 5V or 6V be adequate)?

The TB303 has a custom microcontroller. That will probably be the biggest obstacle to cloning one.

[technix]

Thats funny, buy a 5v regulator and hope for 5.333.
I,m not sure how to connect the Schottky diode.

In one of my designs I needed a 5.25V +/- 5% rail. I put a 1N5819 in series with the feedback pin of the LM2596-5.0 fixed voltage switch mode buck converter chip and got the 5.2V that way.

For 7805, connect the GND pin of the 7805 to the anode of the Schottky diode, and ground the cathode of the diode. This way the diode shifts the reference voltage of the regulator chip up with its voltage drop. This configuration can have bad line and load regulation due to the diode being present on the ground instead of the feedback pin. For (mostly switch mode) regulators that have a separate feedback pin put the diode there to create a voltage drop in the feedback loop.

[mikerj]

I checked the schematic and it uses an 1SZ46A reference diode which is not accurate, but has a very low temperature drift. You will find the datasheet here. It is tweaked with a pot in the schematic, so if it says 5,333V, please tweak it with a 4 digit multimeter to 5,333, but do search for a 50ppm/C reference, because thats the spec on the 1SZ46A.

An LM431 (precision TL431) has 50ppm tempco, seems a good fit for this.

[Crumble]

Hi, i wanto create it from scratch.
[...]

Good luck with that! I like your enthousiasm, but I won't be able to help you with that... I also don't think the 5,333V is going to be your biggest problem.

[DBecker]

Looking at the schematic, this supply is the only tightly regulated one.  It serves as the supply voltage for several critical sections, notably the 7174 flip-flop.  The flip-flop outputs feed the precision resistor network made of R74-R90.  At a glance, that appears to be acting as a DAC, which suggests that stability and noise rejection is more important than the exact value.

Edit: Oh, the block diagram labels this as the D/A converter.

I would substitute an adjustable LDO regulator with a high PSRR.  You might as well get precision resistors to set the voltage, since they will cost barely when buying small quantities.  Or you can modernize the whole section and go with an integrated DAC.

[dmills]

The trap here is that the thing (like most analogue synths) uses 1V per octave for the control voltage, so that voltage is probably fairly critical to making the whole thing stay in tune, including from one octave to the next), but that said, it was a rather simple low cost consumer oriented design and the things always did drift like a pig.

Pick a modern bandgap reference, divide with 0.1% low tempco resistors, buffer with an opamp, I bet you can get close enough without any real need to trim this.

The micro will be a problem, but it is nothing that a modern part combined with a decent programmer could not sort out (Together with adding useful things like USB Pattern load and save and maybe MIDI).

regards, Dan.

[DBecker]

The trap here is that the thing (like most analogue synths) uses 1V per octave for the control voltage, so that voltage is probably fairly critical to making the whole thing stay in tune, including from one octave to the next), but that said, it was a rather simple low cost consumer oriented design and the things always did drift like a pig.

Ahhh, that's the piece I was overlooking.  I thought that the DAC was directly generating a waveform.  Instead it's generating a voltage, which is fed into a VCO to generate the tone.  I now see why the reference voltage is important for tuning.
It's even more critical that the DAC resistors are matched, or the synthesizer won't generate an equal tempered scale (a correctly tuned 'A' will result in other notes being off-frequency).