Relay coil flyback
Yeah, it can be pretty small for some relays. If they have a coil that starts pulling towards the 100s of mA, it’s probably good to look at adding a diode. Even just a simple 1N914 could help.
The need for FET drivers
Forgot to explain why a driver works better then a simple transitor-on resistor-off front end to a FET. The gate of a FET does act is a capacitor as has been noted. The transistor circuit does push a bit of current into the device to get a quick gate charge going to turn on the device. But when it’s shut off, the resistor only slowly bleeds off the charge from the gate, causing a slow turn off which allows the FET into the linear heat dissipation region. This is bad when trying to PWM the device at any significant frequency, as the device may not have completely shut off when another turn on signal comes in spending a fair amount of time warming things up. If you have another transitor actively shunting off the charge, the device turns off much quicker, and doesn’t heat up as much, and all is happy in silicon land. A driver is really a must when building a FET based h-bridge, as the lower transitors on one side must turn off before the upper on the same side turns on. If both are on at the same time, it’s called shoot through, which is a direct short to ground.
Irony
GroG, the word you’re struggling for is “grammar”. LOLs.
Sure sounds like fun. What
Sure sounds like fun. What is the criteria? The largest capacity circuit with the minimal amount of components and cost?
I think it would be great to do something like an open source project like the OSMC- I think it’s a little overbuilt for all our purposes here. Also $169.00 for parts for the OSMC seems pricey.
What should the interface be? Just a PWM input and direction? Fuse protection? A safety relay cutoff?
If we can agree on what we want - it makes designing the thing much easier?
Also what is the output of the design? You need a bridge Chris, and so do I. But, unlike BOA (PCB master) I have never made my own PCB. Do we make art for a PCB or just free form it all together? Do both?
Again, if we address these decisions before we build a skematic, it will save time and expense in the long run.
(Also to be a nag - BOA, is there a way that multiple people can edit a Tip & walkthrough node? - I think we could all benefit if we could start a collaberative doc(s) )
GroG
collaberate design thoughts
For my personal purpose, it would be enough if I could find a summary of Chris’ lessons learned (excusez le management lingo). A list with links and a basic circuit scheme that would explain to me what I need to know, so that I can start making my own decisions and designs for a motor controller that can do more than the standard controllers newbies start out with.
Shorter sentences Rik!
I don’t need a LMR branded PCB. I need to be able to decide:
if I need BJT or FET
if I need PWM or some other control
if I need protection against high currents
if I missed anything
And a bare bones example circuit that would indicate where to add the doodahs I decided to have.
And I discovered that component nodes are editable by all. But not in wiki style (history, restores, discussions, etc.).
Rik
Some h-bridge answers
Ok, car stereo stuff : A wire guage table shows that about an 8 guage (AWG), about 1/8" (1/4" with insulation) can carry 70 A or so. Power transmission would be a little bigger. There are all sorts of discussion about whether “true audio” can be transmitted accurately with a “skin depth” of only xxx, to muddy tech details. What sells is what appeals best to a consumer that thinks “bigger is better” hence “Monster” cable. Would you get someone to buy big expensive wiring and such, if he knew smaller cheap stuff would work just as well?
The TC4427 is a good low side driver, meaning good for a FET being switched to near ground. It works great for the low side FETs used on BOAs relay h-bridge, and could be used in P-channel over N-channel FETs for an h-bridge limited to about 12 volts or so. More complex h-bridges would require more detailed drivers, like the HIP4081A used in the OSMC, or some of IRs devices like the IR2184, that have both low and high side drivers to accomadate using all N-channel FETs, by generating the high side driv.
Motor reversal / flyback diodes go on the outputs of the h-bridge where it connects to the motor. L298 circuits show a good layout for diodes. Essentially a diode is connected from the ground wire to each motor terminal (bar (cathode) to terminal) then anotherattached from motor terminal to the positive voltage (bar to positive). The relay coil would need a diode connected in parallel from the transistor side to the positive side (cathode to positive). Edit to add : in BOAs relay circuit, the “ground” for the diodes to connect to would be the drain side of the FET, instead of the actual ground.
The difference in parts prices and built prices is primarily “overhead”. You, as a carpenter, charge for a service. That paint can only cost you $2, but what do you charge to spray it on something? And performance, and getting the idea that this board built probably already works andif not you can return it, etc. A relay h-bridge requires you to slow to low PWM before reversing, where a solid state h-bridge will operate regardless (up to a point).
Uh, you wouldn’t believe I
Uh, you wouldn’t believe I did that on purpose would you? 
Ok guys, check my math here -I’m ready to order parts…
Well, I found this and it seems to include everything we have been talking about.
I think I will be working with this basic schematic to start building my controller.
If you guys would, below is my list of stuff to order including in bold the items you guys have suggested and where I am going to substitute for the parts listed above. If you could check my math here, I would love you guys for ever and ever. (Assume I am doubling all of this as I will be driving 2 motors)
Instead of the IRLZ44's or IRF3xxx's I am using (4) IRF2804
Instead of the (4) 1N4005 (D1-D4) I am using the MBR20150CT **I still don't understand the pin-out of this unit. It shows both diodes shooting to eachother meeting on the middle pin. It seems that I need them both facing the same direction as shown above. Do I need a total of 4 -using only 2 pins of each?
Instead of the National FET Driver DS0026 I am using a TC4427 **Now, I am led to believe through what I have read and what is on the schematic above, that the IRF2804 is a N-Channel and needs the gate to be pulled to ground to open the source to the drain. If this is true, and the PWM comming out of my Picaxe is a + logic signal, do I need to get the TC4426 instead of the TC4427 as it is an "inverting" unit? Again, I am led to believe that if V+ Logic is comming in and I need to switch ground on and off to run the IRF2804, I would need to invert. --I dunno, just tell me which one to get.
The rest of the stuff shown above seems to all be self-explainitory, caps and the smaller flyback diodes going around the FET's themselves. --I think I got that figured out and covered.
And just one more: I would assume that (as shown above) PWM input A and C should be tied together and PWM input B and D should be tied together. No?
--That should do it! I can't tell you guys how much I appreciate how much help you have been on this and all my other projects, Thanks. Like I said above, just let me know if I am on the right track and I will get this order, well, ordered!! --I can not wait to start etching the board!!!!!!!
A few h-bridge issues
IRF2804 is a good FET, should work weel.
The MBR20250CT part I glossed over and should have been thinking about better. It is a common cathode part which is good for the top side only, connecting each motor terminal (with pins 1 and 3) to the positve voltage (pin 2). What is needed is a similarly rated common anode 20 A schottcky barrier rectifier to go from the ground to each motor terminal. This is mentioned in another message above, in answer to some questions earlier on connected the diodes.
The TC4427 is a good part for low side drive on N-channel FETs, it won’t drive the upper FETs in this h-bridge. It is a great part to use for BOAs relay h-bridge since both FETs there are low side, sources to ground, but not here. It might be workable with a P-channel over N-channel h-bridge, limiting it to 12 volt power only. For an all N-channel FET h-bridge, perhaps an IRF2184 or HIP4081A would work as a driver with both high side and low side capability. An N-channel FETs gate has to be switched 10-12 volts higher, not lower than it’s source. P-channel FETs are switched to a source 10-12 volt lower than the source.
I’m not sure how well the above h-bridge graphic will work, the inputs are a bit screwy with the 10 uF caps and diodes on the gates. The output placement of diodes is good, but the diodes used should be the higher speed schottky rectifiers.
I was on the mailing list for the development of the OSMC, and it took a few months to get a consensus on parts, calculations etc. This should be simpler, and should only take a few days. It’ll happen, just not good to rush.
4x N-channels?
Can any of the Uber-Boffins comment on the use of Rb and Rt? Presumably that’s how the gate voltages are regulated after the charge pumps. Can this work? What would the PWM Vpk-pk be?
How do you DO it?
…manage to compose the answer at exactly the same time I’m composing the question?
High side and Low Side FET’s
Ok, robotologist, I think I got this high side low side stuff… The high side is switching the + on and off and the low is ground. If indeed I need a half and half driver there is a TC4428 which is half straight and half inverted. --Probably what I should be looking at. Or maybe I should go back to considering the relay idea. I swear, the more I learn about this stuff the more I find I don’t know.
Oh, and one request… When you guys suggest a part i.e. schottky rectifier or whatever, would you be so kind as to include a googleable part number? --It REALLY helps me a lot when I can find a data sheet that I can study. Remember folks, you have to give me my homework to do.
No, wait a sec…
high side, low side… Now I am really confused… there is no positive/negitive on the gate… isn’t a n-channel negitive to open and p-channel positive to open? Damnit! I SWEAR I am going to figure this out! --Well, I do love puzzels!
Not sure on the resistors
The front end to the FETs seems very strange in that diagram, and usually resistors, particular 100 k ones, only serve to slow turn off times configured as shown. I glanced at the part that is to be attached to it, but haven’t any clues from that.
H-bridges
What I’d like to try is putting together schematics and parts lists for 3 different h-bridges. In the posts, I’ve mixed a couple parts up, and have noticed different parts where they shouldn’t be. Like the schottky diodes, they really should be for the relay h-bridge, and not necessarily for a FET h-bridge, and need a matching set of common anode diodes needed too. And the low side driver also for a FET under relay h-bridge, but could be used on a P over N h-bridge too.
BOAs relay h-bridge is a good idea, and workable, and just a few parts could be added to it. That would be a first possible draw-up.
Putting together a P over N h-bridge using the TC4427 (or close relative) would be another possible design, though it would probably be limited to 12 volts.
And finally drawing up an all N-channel FET h-bridge based on an actual high and low side driver (IR2184 (not IRF) or an HIP4080 or HIP4081)
Will need a bit for parts lists, but does this sound reasonable?
Oh, some great information on learning about h-bridge is found in an application note for the LMD18200
I’m right there with you…
I’m right there with you Chris, I am so confused on some of these topics on motor control. Alas I’m even further behind, I’m ordering my PICAXE tomorrow…/sigh
I had a question on the posted schematic, why are there 4 PWM inputs on that H-Bridge?
@ Robologist…
You are a gental man and a scholar. And I am a patient guy. I would love any draw-ups and part lists you got, and am happy to wait for them. I do like the idea of a few different versions --gives me (everybody) more to learn from. As for me personally, I will not be exceeding 12V but would love to see them all. Thanks bunches.
Oh, I gotta finish with a question…
Tell me again the name of the company/ website with all the componants? --The one where you can order the HUGE catalog… Was it national semiconductor? I can’t remember.
Clicky H-bridge
BOA’s relay H-bridge is not only workable: it’s a reality. I came up with it after I’d been down all the roads CtC is currently exploring, having decided they were all too complicated or too expensive!
Chris, I tried to make your life easy, but you have chosen the difficult path, my child. I hope you fare better than I.
Hello All,I am working on
Hello All,
I am working on schematics for multiple sizes of H-bridge … might as well cover them all in one spot (I know I would benefit from it) - from the very small 2n2222 to the very large FET (possibly parallel FET)
There will be PROS and CONS, max voltage and amperage ratings, maximum PWM frequency, List of parts, possibly Art work and Gerbers at some point… Ooooh the thrills … the chills … I’m tingly all over !
Each will be layed out in a standardized format in one of the Edit by All component nodes, rik so gratefully brought to my attention. That way BOA, robologist, Jip, JKA, rik, & CtC can correct my diagram and grammer directly. I will put a copy of the whole thing on my wiki so we can have revision control, since I don’t believe Drupal does this.
I’m so excited and giggly… !
GroG