Shop OBEX P1 Docs P2 Docs Learn Events
Lon Glazner's High Power Bridge — Parallax Forums

Lon Glazner's High Power Bridge

JimGJimG Posts: 84
edited 2011-09-25 20:06 in BASIC Stamp
Lon did and article in Nuts adn Volts in conjunction with Parallax titled Stamp-controlled High-power H-bridge. It was Column #52, in teh August 99 issue. Can anyone help me with this circuit a littl ebit? I realize it is old but I am interested in it. I hae it build but it doesn't work. I AM NOT using the Stamp with it yet. I am just connecting the appropriate pins to the + or - side of the chip power supply to simulate the Stamp. I have attached the article.

Thanks,

Jim
«1

Comments

  • Mike GreenMike Green Posts: 23,101
    edited 2011-06-11 20:04
    You must draw a schematic of exactly what you've built and attach that to a message (not what you think you've built ... trace it out). If what you have doesn't work, it's not what's in the Nuts and Volts Column. You have to be very explicit about what you've tried, what you've measured, etc. It's useless to tell us that you've built what's in the Column and it doesn't work. We need details about what you've actually built and how it behaves.
  • JimGJimG Posts: 84
    edited 2011-06-11 20:22
    Mike. Well, my initial question was more along the lines of whether or not anyone had any interest in helping with a circuit this old. I fully realize that what I have built is not what was in the Nuts and Volts article because a) I am not driving this with the Stamp yet, and b) I am using a couple of bench power supplies for power rather than the power supply circuit shown in the article. Beyond that, it is that same. I have attached a schematic of EXACTLY what I have built. I have traced it and retraced it.

    Jim
  • Mike GreenMike Green Posts: 23,101
    edited 2011-06-11 21:10
    Looking at your wiring and at the truth table in the HIP4081A datasheet, it looks like the low transistors should both be on and the high transistors off. I recommend looking at the truth table. I also suggest that you have some switches or jumpers for the HIP4081A inputs so you can change them easily to see what happens with different settings of ALI, BLI, AHI, BHI, and DIS.

    You still haven't described what happened that you describe as "doesn't work".
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-11 21:14
    You need to have (it's either):
    1) AHO[Q4] and BLO[Q3] on
    or
    2) BHO[Q1] and ALO[Q4] on.

    The objective is Q2 and Q3 for one direction, Q1 and Q4 for the opposite.

    Your H-Bridge.pdf shows you with ALI and BLI hardwired to + and BHI and AHI hardwired to GND.
    You need to have AHI and BLI on or BHI and ALI on.

    If the A highside is on, it needs to have the B lowside on, too.
    Conversely, with the B highside on, the A lowside should be on with it.
  • JimGJimG Posts: 84
    edited 2011-06-11 22:10
    OK, I made some changes. Still doesn't work. I have attached a modified schematic with some voltages at various points. I think I have looked at this too long, I am not seeing it anymore.

    Jim
  • Mike GreenMike Green Posts: 23,101
    edited 2011-06-11 22:17
    If you really want help, you have to pay attention to the suggestions that people make and show in some fashion that you're paying attention.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-11 22:18
    Your PDF cited has Vg (11V) for Q1 and Q2.
    Not in conformance.

    PE -- The IC's outputs aren't right.

    I'd disconnect the FETs from the IC till I got those IC outputs to come right.

    Having had both A-side FETs on together (or both B-side) didn't do them a world of good either.
  • JimGJimG Posts: 84
    edited 2011-06-11 22:33
    Mike,

    I am not sure what you are referring to?

    JIm
  • JimGJimG Posts: 84
    edited 2011-06-11 22:46
    I have uploaded a new troubleshooting schematic. I am not sure what you mean by PE in one sentence. The Voltages on teh FET's are much different since I changed the connections on the driver chip. I also modified the schematic to indicate that I am using .22mf ceramic caps rather than electrolytics, not sure if that is an issue.

    Jim
  • JimGJimG Posts: 84
    edited 2011-06-11 22:49
    Do you think I trashed the FETs? I have more. If need be, I can swap them. I can also disconnect the IOC fromt FET's easy enough too but am not sure how to test to see if the ICF turn turning on the right lines in that case.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-11 23:02
    "PE" means Post Edit
    "IOC", "IOF", now you've got me. Not important.

    You need, as I see it, to make sure that the IC outputs are appropriate, given the respective inputs. So, "half-split" it, take the FET circuit out of the mix.

    Get out the voltmeter
    [I could suggest using LEDs and resistors. Would that be opening another can of worms?]

    With ALI and BHI high & BLI and AHI low, then ALO and BHO should be high out and BLO and AHO should be low.
    Agreed?

    (BLO and AHO -- L-O-L!)
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-12 06:17
    So, this HIP4081 (which is not a CD4081) is supposed to prevent bad combinations of the control inputs from resulting disastrous FET conditions.
    The test results noted indicate... non-compliance.
    I don't know your skill set or how you have this all laid out.
    I hope you don't have all of the FET tabs mounted to a common heatsink.
    It'd probably help if you placed a "map" on the table that you can refer to when you're sorting this out.

    (I would've used different designators for the transistors which would reinforce the pairs' harmony and not the author's clockwise-arbitrary method.)

    keepcalm.gif
  • JimGJimG Posts: 84
    edited 2011-06-12 09:42
    I am working on your other post. I like the LED idea. I will hook them up and see what I get. This is a great idea.

    The "IOC" was a typo, I meant IC.

    My skill set is marginal, at best. I am generally a pretty good adapter and modifier but not much good at ground-up design. My interest in all this is drivers for robotics project I have going.

    I do not have the FETs on any heat sinks yet and I have them staggered so that the drains are not touching. I will heat sink them once I get the basic circuit working. I am testing with a very light weight motor, maybe 1A, so I should not have a heat problem.

    I am open to your suggestion as to designators. Anything the helps and improves my knowledge base is greatly appreciated.

    I am uploading a couple pictures of my breadboard, maybe that will help you.

    Thanks,

    Jim
    511 x 599 - 238K
    1024 x 768 - 106K
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-12 10:05
    There's a tendency to scrunch stuff together. Life will be easier if you spread stuff out, you can get in there with your voltmeter probe without worry that way.

    The highside transistors provide alternate +V sources and the lowside transistors provide alternate paths to ground.

    One way it's Q2/AHO---(a)motor(b)---Q3/BLO (Q1=off,Q4=off)
    The other way it's Q1/BHO---(b)motor(a)---Q4/ALO (Q2=off,Q3=off)
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-12 10:15
    It's covered in the article, but I've attached a color-accented diagram with some notes
    covering the objectives.
    483 x 490 - 39K
  • JimGJimG Posts: 84
    edited 2011-06-12 10:20
    Thank you for both replies, I will fiddle some more and get back to you.

    I did get 4 LEDs hooked up and two light but I haven't gone any further with that yet. I haven't had time to figure out which two yet and I haven't tried ther alternate postions.

    Jim
  • JimGJimG Posts: 84
    edited 2011-06-13 20:52
    Ok, here's where I am. First, I substituted LEDs for the FETs as shown in the attachement. Yes, I do still have the stray connections to the FETs and guessed it wouldn't matter. I have attached another schematic showing what this looks like. With this setup I get PIN 11 AHO and PIN 18 BLO high with the corresponding LEDs lit. I get PIN 13 ALO and PIN 20 BHO low with their LEDs dark. If the reverse connections simulating reversing the motor I would think tha tthe opposite set of LEDs would light but I get nothing lit. Am I giving you enough info?

    Jim
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-14 06:11
    Your pdf/dwg doesn't show the LEDs with series resistors.
    Maybe you forgot to show them?

    From your data, the BLO BHO output is wrong as it's approx 9V when it ought to be 0.

    PE - The IC's mnemonics leave much to be desired.
    544 x 528 - 42K
  • JimGJimG Posts: 84
    edited 2011-06-14 10:02
    You're right. I was hurrying trying to get the reply up to you. I will fix it and repost it.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-14 12:13
    Hurry? What's the rush?
    The article is almost 12 years old -- :)

    Had a big-wig used to tell us "Speed wins". He was wrong.
    Slow and steady progress, accurate info, get it done right.
  • JimGJimG Posts: 84
    edited 2011-06-14 15:09
    I always feel like you guys are just sitting there waiting for my missives...that how important I am. :-)

    I have to do some fiddling then I will get back to you on progress. I think one thing I need to do it use a much bigger resistro as I think I am drawing too much power of the pin in the IC. What do you think?
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-14 16:46
    a much bigger resistro as I think I am drawing too much power of the pin in the IC. What do you think?

    What resistor where?
    I'm not busting your chops, but you need to be specific.

    Use insulation stripped from wires to cover resistor leads and so on.

    It stands to reason that, knowing nothing else, if you get the wiring right then it should work.
    [That best laid schemes o' mice and men gang aft aglee.]

    Do you get/see my point about those output states vs their inputs?
  • JimGJimG Posts: 84
    edited 2011-06-15 09:03
    Resistors on the LEDs.

    I agree, it ~should~ work just dandy if the wiring is correct and the parts are the same. There are two parts that I used that are different. The zeners were not specified by part number so I chose a 15V zener adn the caps are ceramic rather than electrolytic.

    If you mean the relationship bteween the inputs and outs nput, yes, I get that.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-15 11:36
    The zeners limit the gate voltage. What you've done is likely acceptable.
    Ceramic vs electrolytic, probably doesn't make a difference in this case.

    OK, so your work is cut out for you. :)
  • JimGJimG Posts: 84
    edited 2011-06-15 12:25
    I didn't need more work, I'm already married. This was supposed to be fun. :-)
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-17 15:53
    I made half of an H-bridge, just for grins and giggles.
    Take a look at the schematic attached.
    I made some notes.

    I used IRLZ14; like Glazner's, their gate threshold is "logic level".
    I disagree with Glazner's not using commutation (suppressor, free-wheeling, et al.) diodes across the FETs. I did use them (D1, D2.)
    I didn't use Glazner's zener regulators either, just pull-down resistors.

    When Q2's gate voltage > threshold, it's on, its drain will be nearly at source potential (ground). Increasing V_gate past threshold will result decreasing V_DS (better conduction, bringing drain closer to ground).
    Q1, the "high side driver" is configured as a source follower. The source voltage will be less than, but will "track", the gate voltage.
    Q1 won't be in full conduction, so V_supply - Q1_V_source is across it (Q1 V_DS.)
    So, with a 15V supply, I got 11.5V across my motor, running about 50mA (no load) and about 3V across Q1.
    678 x 378 - 20K
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-06-21 09:19
    Thought somebody would've asked about "bootstrapping" before now.
    Oh well.
  • JimGJimG Posts: 84
    edited 2011-09-25 19:26
    PJ,

    I didn't see these until tonight. I am looking at your circuit and the first question I have is how does it work WITHOUT bootstrapping? I the the gate voltage had to be at least V_gate plus the motor supply voltage, am I nuts?

    Jim
  • JimGJimG Posts: 84
    edited 2011-09-25 19:40
    PJ,

    According to the datasheet the FETs are diode protected internally, or do I misunderstand that too.

    Jim
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2011-09-25 20:06
    I took up the cause and started a subject in another forum --
    http://forums.parallax.com/showthread.php?132721-H-Bridge-Development
Sign In or Register to comment.