auto radio solid state vibrator for vintage tube radio

[sphinx]

i found this on YT and it is a replacement for a mechanical vibrator for creating high voltage on vintage car tube radios

i found this interesting and a fun small project

for some reason he made this private but i succeded to download before he did that

please check schematic for errors it is video please watch the pin on mosfet seems to be right for the irf3205
some info in schematic sheet

adjust for your own preferences this can easily be made with smd as well if wanted.

url file is for the video 

[sadman]

please upload schematic in pdf format
 

[sphinx]

here you go, please be advised to check with what is shown in video/

/sphinx

p.s. i forgot to say that the users name is shango066 that made this on his channel
next time i will try to remember to create a pdf file for shematics for those that might want it not much extra work for that thanks at @sadman

[sphinx]

i found a missing trace on pcb between pin 6 and 7 on IC yoou can correct it or download a corrected

i checked schematics no errors

the schematics is still same no edits i just made it a bit more readable there just the ocb that got a missing trace.

pdf is included as well

sorry for the mess

/sphinx

p.s. from now on i will not make the schematic designs compact it alot better to have easy readability

I MADE A NONO ON THIS ONE DELETING FILE WILL REUPLOAD WHEN I SORT THINGS OUT

THERE SHOULD NOT BE A CONNECTIONS BETWEEN PIN 6 AND 7 THAT WILL CREATE A SHORT BETWEEN 12+ AND GROUND
THERE WAS MADE AN ERROR WHEN MOVING COMPOENTS

THIS A NEW REUPLOAD AND THIS ERROR FIXED

I ADDED ALLREADY MADE GERBER FILES NO OTHER CHANGES MADE TO DESIGN

[PM3295]

I watched some of shango's videos in the past. Interesting guy.  The proposed circuit works ok, but there is no duty cycle control, if you need to reduce the final output voltage. I have done a design in the past (also single chip) that serves the same purpose, but has some duty cycle control as well. If someone is interested, I can dig out the diagram and post it here.

[sphinx]

it is mentioend in the video that voltage can lowered with diodes on the common source line from both transistor, he had some problems with that he talked about it

why not share it i could make a pcb and gerbers for it it seems llike a nice idea this could be use for other purposes as well.

[PM3295]

It was designed for driving a push-pull stage in open loop, with duty cycle control. I will look for it and post the schematic.

[PM3295]

I couldn't find my original design, so I had to make one up from memory, breadboard and test on the bench to be sure. You can experiment with the adjustable legs if you want more variation for duty or frequency. Out A/B is the drive to the MOSFETS.

With the actual circuit, the duty adjusts between zero and 49%, but in the Proteus model, the same values only adjust down to about 30%.

[sphinx]

could you please share the proteuss projest so i wont need to redraw that design

[PM3295]

Sure!

Posted on: January 31, 2023, 11:00:12 23:00 - Automerged
There is something wrong with the UC3525 model. With the real circuit, adjusting the pot to min is also min duty, as the chip should function. In Proteus, it is reversed. So, something is up with the model. It looks like the compensation pin function is not modeled correctly.

[sphinx]

is this same design for a transformator with middle tap and can i use 2 n type mosfets as with my old design ?

i tested it in proteus and it works just fine only thing i can see is thatso called "duty cycle" doesnt go up to edge to edge on the pulses
i guess it will work just fine as it is, now output voltage can be adjusted to the liking.

[PM3295]

Yes, it is meant for push-pull with center tapped transformer. In the real circuit, the duty cycle adjust over full range. It is the model messing around in Proteus. I checked the chip operation in Cadence Orcad, and their model works correctly as far as the compensation pin.

[sphinx]

i am almost finished with the schematic so i will soon create a pcb

[PM3295]

If you want to add some dead-time, you can put a 100 ohm or so in series with pin 7.

[sphinx]

i need to look up what that means for the chip and signals. thnx for that extra input

[PM3295]

In my original design, I also used pin 10 for current sensing and protection. If you want to implement that as well, let me know.

[sphinx]

man when i just completed the you just had to come up with something new so i need to re route the design, just kidding

that could be a good idea to have some protection of the circuit

i attach the new design a work in progress with pdf if you want to have a peek

just wondering if i really need to concerned about component placement that much since this will be 80-200Hz ?

[PM3295]

Maybe, you can design for two variants? One simple without the added features, and one with the current sensing protection. The last one we need to add another low-value power resistor. I will make the changes on the schematic for you.

Such low frequency should not be critical for layout.

[sphinx]

well if only a resistor is to be added for shutdown then it is ok it can be added to this design like the resistor on pin 7 it could be bridged if that function is not used

i checked the data sheet not much info on the shutdown from what i saw

ok roger that on placement that was what i was thinking.

[PM3295]

The value of the power resistor (0.1 in diagram) must be such; that peak current (under normal conditions) through the FETs generates a peak voltage < 0.7 V. The RC filter value will depend on the transformer, leakage inductance, etc. The time constant should be just long enough to remove any initial short spike, that may trigger the current limit prematurely. So you can play around with the R while scoping the waveform on pin 10 for sort of a saw-tooth signal, without any starting spikes...

[sphinx]

thanx for this addon i will add these to the design, these can be left out as well if one doesnt want this function

[PM3295]

You also want to minimize any voltage drop caused by switching current between the points marked, when using the current sense. Try and group those points together. Otherwise the bounce in the ground line will interfere with the current sensing.

[sphinx]

this is the design with the latest modifications with current sensing have a look at it.

[PM3295]

I will try and route it like this to have equal lengths of track on each source connection. Best to keep it sort of balanced for push-pull.

BTW, you have an error on the gate connection T2.Recheck against my diagram.

I see more errors Vref. Diagram should look like my attachment.

[sphinx]

roger that, will fix it and add documentation/info for those extra components i will also widen the power traces as well

now this works for both the first versio and with the extra additions.


Posted on: 2023-02-01, 05:04:02 - Automerged
this is with the fixed traces

will now sort out the component labels and documentation/info

will now take a cup of tea and watch some crime tvshows before bedtime.

will fix the stuff la

Posted on: 2023-02-01, 05:20:29 - Automerged
i fixed the t2 error with the gate and vref error those were easy to fix on pcb as well

sometimes altium does funny things if not careful when moving lines or components

now i just going to add info on schematic if someone wants those extra features or not but that will happen after i sleep.

brain is getting a bit woozy so better to stop now and do the rest with a fresh brain.

here is the latest fix for those errors i made.