# ULTRA FAST H-BRIDGE HELP...

Discussion in 'The Projects Forum' started by franzschluter, Jun 3, 2009.

1. ### franzschluter Thread Starter Active Member

Jun 1, 2009
95
0
Hi guys,

I need advice in building an ultra fast H-bridge.. Designed to switch in the preferably 10us range? Is this possible? I used SN754410 from Texas but it doesn't seem to do the trick.

Application:
I got an Homebuilt EDM ram that's controlled via a hydraulic system.. I tried using a LM339 circuit that measures voltage differences to make the ram maintain a certain voltage threshold window or gap.. Meaning Open circuit is 100V and Short Circuit is 0V. Let's say I want a spark gap of 75V. My comparator will compare and try to maintain the gap between the positive and negative terminal by driving the ram in a certain position such that 75V is there.

Problem:
I think my comparator is not fast enough to maintain the ram. Since it is vibrating a lot. Optimum would be it'll be switching my H-bridge so fast that it'll practically maintain a gap of 75V and not vibrate that much. In my case here.. It is swinging up and down in a very slow manner that i can see with the naked eye?

Question:
What H-bridge setup may I use in order to make up and down movement respond faster... The hydraulic movement up and down is controlled by my h-bridge.. The movement is controlled via 12V or 36V in the 600mA range. Reversing voltages makes ram go up and down.

2. ### bertus Administrator

Apr 5, 2008
15,799
2,386
Hello,

Can you place a schematic of the circuit you have upto now?
It can be that the LM339 is not fast enough.

Greetings,
Bertus

3. ### franzschluter Thread Starter Active Member

Jun 1, 2009
95
0
LM339 - Low Power offset Voltage Comparator
Pins..
1 output 1
2 output 2
3 v+
4 input 1 -
5 input 1 +
6 input 2 -
7 input 2 +
8 input 3-
9 input 3 +
10 input 4 -
11 input 4 +
12 v-
13 output 4
14 output 3
SN754110 Quadruple Half-H Driver
1 1,2EN
2 1A
3 1Y
4 NC
5 NC
6 2Y
7 2A
8 VCC2
9 3,4EN
10 3A
11 3Y
12 NC
13 NC
14 4Y
15 4A
16 VCC1
There are two potentiometers in the circuit 10k.. This is to set the threshold level. I used voltage divider from the 100V line such that the comparator can compare it to the 12V as reference.
The right potentiometer sets the reference of the 100V and comparator.. The left potentiometer sets the 12V reference..

My device works.. In general... But my ram is vibrating too much...

In theory.. The comparator will compare or must compare the voltages fast enough that it will be able to decide if the ram must move more up or down to reach a specific preset voltage or "spark gap". The comparator sends its decision to the H-bridge and the H-bridge will tell the ram go up or down. The ram up and down is controlled via a precision hydraulic solenoid that depending on the polarity it will go up/down.. 12V up to 36V can be placed in the coil.. I tried increase hydraulic pressure and voltage for solenoid (Hoping that it will move/respond faster to make the vibrations less noticeable but no luck it just got from bad to worse). My doubt that comparator and h-bridge is too slow.. I do not know which is slower? Or why the process of the circuit is this slow...

I am afraid of using mosfets as h-bridge because it might short out due to the very fast switching needed.. But a mosfet bridge would definitely be faster (only my opinion). What I need is a very fast H-bridge circuit and comparator that doesn't need to drive a high load, 1A is more than enough for this solenoid coil.

File size:
15 KB
Views:
67
4. ### bertus Administrator

Apr 5, 2008
15,799
2,386
Hello,

Pin 1 and pin 2 of the LM339 must not be connected to ground.
The outputs of the LM339 are open collector.
You will need a resistor on the outputs.
The "optimal" current for a comparator is about 6 mA.
The resistors you have taken are to large.
Try 2K2 resistors for R5 and R8.
This will probably make the output signals of the comparator more stable.

Greetings,
Bertus

5. ### franzschluter Thread Starter Active Member

Jun 1, 2009
95
0
Thanks a lot Bertus for advice. I will try replacing it with 2k2 and see what happens. My R5 and R8 is pull-up resistor only. So that it is normally on the high state but will go low when told by comparator...

You also mentioned pin 1 and pin 2 must not be connected to ground so this means I'll put 10k ohm to ground or VCC? I assume VCC because it is collector correct?

For optimal current at 6mA you mean it's input or output? I guess output correct?

Everything you mentioned makes sense now.. Maybe this will give greater impact..Fingers crossed

Best regards von
Franz