-1
\$\begingroup\$

hi i am working on the design for a device for my master thesis and i want to use two voltage transformers small as siye of quarter finger and i want to ask i used the output of first transformer as input for second transformer my ratio for each is 10 and i wanted to get 100 times the voltage the circuit is supplied by two pushbull transistors on ltspice simulation it worked fine but on real life not working any idea maybe the connecting of two transformers need add or sth any help is appreciated i am stuck here since two days thanks enter image description here

output after only one transformer on the oscilliscope give me 10 times magnifing for the voltage on the left and the output after two transformer on the right doesnt give me what i want (the blue always is the output after pushbull and the orange is the output after transformer/s) enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ Your grounds are messed up. \$\endgroup\$ – Ignacio Vazquez-Abrams Aug 8 '16 at 13:04
  • 1
    \$\begingroup\$ the ground is just for the ltspice programming and i could do the transformer in parallel and got 2*10 times and work like a charm also the ltspice work and give me 100 times although its the same what i have built on ltspice! thanks \$\endgroup\$ – Abdallah Mteir Aug 8 '16 at 13:15
  • 1
    \$\begingroup\$ Have you tried measuring the voltage across M4? \$\endgroup\$ – Ignacio Vazquez-Abrams Aug 8 '16 at 14:21
  • 1
    \$\begingroup\$ Start simple. Only use one transformer. Does it work? Maybe? Almost? Not at all? Maybe you could post pictures of your oscilloscope traces? Do you have a scope? \$\endgroup\$ – WhatRoughBeast Aug 8 '16 at 15:27
  • 3
    \$\begingroup\$ @Abdallah Mteir: Is your keyboard possibly missing the SHIFT key and keys for punctuation marks or why don't you use them? \$\endgroup\$ – Curd Sep 8 '16 at 13:30
1
\$\begingroup\$

Think about the loading effect of the 2nd transformer's primary on the first transformer. Impedances are magnified or diminished by the turns ratio squared. Your 1st transformer is loaded on its secondary by 10 mH and this is reflected down to the transistor output by a factor of 100 (1st transformer's turn ratio squared).

It now appears as 0.1 mH as a load for your transistors. If you designed your driver to cope with a 10mH primary impedance you have now swamped this by connecting the second transformer and the transistors now see 100 uH.

That's the first potential problem.

The second potential problem is that you may indeed get 10:1 step up from the 1st stage but the size of that secondary voltage is likely to saturate the core of the 2nd transformer and you'll get very little output.

Try posting a link to the data sheet (not a buying/ebay page).

\$\endgroup\$
  • \$\begingroup\$ thank you sir for replying i tried today to show you what i am getting on the oscilliscope after one transformer and after two transformers i attached them in images above \$\endgroup\$ – Abdallah Mteir Aug 9 '16 at 8:18
  • \$\begingroup\$ here is the data sheet for transistors also for the transformers sadly for transformer not much info available drive.google.com/file/d/0B191MPL34ekwb3hGZld1TzM0R0U/… \$\endgroup\$ – Abdallah Mteir Aug 9 '16 at 8:37
  • 1
    \$\begingroup\$ The scope pictures don't tell me much because I don't understand the voltage scales. Neither can I tell if the left screen scope shot is with the second transformer loading the 1st transformer output. Your link to the transformer data sheet doesn't work either. \$\endgroup\$ – Andy aka Aug 9 '16 at 8:58
  • \$\begingroup\$ the scale for the voltage is should be squared but with more current but for my application is enough to get my signal to the ultrasonic sensor the left is with one transformer right with two transformers the data sheet link is here sir dropbox.com/s/pzjanlzkqcogii5/data%20sheets%20.rar?dl=0 \$\endgroup\$ – Abdallah Mteir Aug 9 '16 at 9:42
  • 1
    \$\begingroup\$ I'm sure these will be useless for cascading (as most transformers are). Also, the data sheet gives no indication of what maximum input voltages can be used before core saturation becomes a problem. Not a useful data sheet. \$\endgroup\$ – Andy aka Aug 11 '16 at 13:08
1
\$\begingroup\$

The usual way to deal with underwater piezo or piezocomposite transducers is to use a tuned matching network.

Look at your transducers G/B plot and you will see that the thing has a fairly large fixed capacitance, you can resonate this with a carefully designed transformer secondary, then do an L network on the primary side to get a drive impedance suitable for your signal source (You might want some series R here to increase the effective bandwidth).

The transformers are usually custom windings on a pot core or similar, and you often need kapton tape between the layers to avoid flashover.

You generally also find that if you can run the H bridge from a hundred V or so the impedances become much more reasonable, a boost converter is often indicated.

My last low voltage design put out 1KVA between 10 and 20KHz from a 12V source, not doing that again, the transformer primary was 1.5 turns of litz wire and the magnetising current was of the order of 100A!

My guess at your problem would be some mixture of core saturation and reflected impedance, but your spice model is horrible (How does the bottom half of the H bridge switch??).

You might find that producing a model of the transducer (It looks like a crystal resonator of rather low Q), and hooking up your transformers with a simple voltage source as a driver, then plotting primary current against frequency instructive.

\$\endgroup\$
  • \$\begingroup\$ Also, 22K reflected thru 100^2:1 is 2.2ohms at the primary of the first transformer so the load on the H bridge is 100uH || 2.2 ohms, a big ask at ~50KHz. \$\endgroup\$ – Dan Mills Oct 8 '16 at 12:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.