H bridge l298n in Parallel Burning Out After Some Time

Thread Starter

rodolfoviolac

Joined Dec 3, 2019
4
Hi Guys,

I have a college project where I need to controll a motor (27V / 3 A), I used a H Bridge l298n in parallel to support the current, It worked for while and then stop but I don't know what was wrong. Does any body know anthing that could help me?

The motor is a Maxon 118752 precision brush motor (20 Watts) operating at 27 V max. The motor offers a peak current of 3 A, with a maximum continuous current of 1.2 A.

Spec for the motor:
https://www.made-for-science.com/de...er-2-dof-serial-flexible-joint-usermanual.pdf

My circuit Layout:
Screen Shot 2019-12-03 at 8.34.41 AM.png
 

Tonyr1084

Joined Sep 24, 2015
5,239
On page 7/13 there are some suggestions. You might want to consider them. This (below) is a data sheet for the L298N chip. May not necessarily be the same supplier as the one you're using. Specs will vary from manufacturer to manufacturer. This sheet suggests 4 amp capability, but I think I saw a different sheet that suggested 2 amps. So manufacturer is important. So is having the right data sheet.

https://www.st.com/resource/en/datasheet/l298.pdf
 

crutschow

Joined Mar 14, 2008
25,663
Is the l298n on a heat-sink?
It likely needs one.
Its total saturation voltage of the two outputs in parallel at 1.2A is about 2.2V, so its dissipation will be about 2.6W at 1.2A.
The device has a 35°C/W thermal resistance to air, so that would make its temperature about 91° above ambient.
 

Thread Starter

rodolfoviolac

Joined Dec 3, 2019
4
Is the l298n on a heat-sink?
It likely needs one.
Its total saturation voltage of the two outputs in parallel at 1.2A is about 2.2V, so its dissipation will be about 2.6W at 1.2A.
The device has a 35°C/W thermal resistance to air, so that would make its temperature about 91° above ambient.
Yes, I do have heatsink on it
 

Alec_t

Joined Sep 17, 2013
11,744
Welcome to AAC!
ST's L298 datasheet shows fast recovery diodes (trr ≤200nS) for spike suppression, but you are using 1N4007 (D1-D4). That might be the problem.
 

ronsimpson

Joined Oct 7, 2019
1,020
I do not know if this is the problem but:
3. APPLICATIONS Fig 6 shows a bidirectional DC motor control Schematic Diagram for which only one bridge is needed. The external bridge of diodes D1 to D4 is made by four fast recovery elements (trr ≤ 200 nsec) that must be chosen of a VF as low as possible at the worst case of the load current.
I do not like the 1N4007! It is not fast. And the Voltage Forward is high. If I had to use a slow diode at least use 1N4001 for a better Vf. You need a "switching" diode or fast recovery diode. I would use a Schottky.
 

Thread Starter

rodolfoviolac

Joined Dec 3, 2019
4
@ronsimpson thank you for your support, can you help me dimensionate this diode?
I have these schottky available: SB360, SB3200, SB5100E, SB560, SB5200, SB3100, 1N5819, 1N5822. What fit the best for this case?
 

crutschow

Joined Mar 14, 2008
25,663
How fast is the motor being turned on and off?
Do you have a pause to allow the motor to stop before you reverse the motor?
 

ronsimpson

Joined Oct 7, 2019
1,020
SB360 3A 60V DO-201 larger case.
SB5100 5A 100V DO-201
1N5819 1A 40V DO-204 I think this is the same size as 1N400x.
1N5822 3A 40V DO-201
With out looking up every part, the 1A diodes are in the small case, 3A in a bigger case and 5A in a bigger yet.

I would try the 1N5819 and see if that helps! If this is the problem then you will know soon.
 

TeeKay6

Joined Apr 20, 2019
572
Hi Guys,

I have a college project where I need to controll a motor (27V / 3 A), I used a H Bridge l298n in parallel to support the current, It worked for while and then stop but I don't know what was wrong. Does any body know anthing that could help me?

The motor is a Maxon 118752 precision brush motor (20 Watts) operating at 27 V max. The motor offers a peak current of 3 A, with a maximum continuous current of 1.2 A.

Spec for the motor:
https://www.made-for-science.com/de...er-2-dof-serial-flexible-joint-usermanual.pdf

My circuit Layout:
View attachment 193638
I question the wisdom of the datasheet statement "For higher currents, outputs can be paralleled." Each H-bridge in the L298 has its own delays and saturation voltages, and those delays and voltages are not exactly the same for each bridge. Thus it is likely that for some (short) portion of each cycle one bridge is driving (e.g.) Out1 high while the other bridge is still driving (e.g.) Out4 low. That is, for some portion of each cycle the device may be fighting itself rather than driving the external load. Also, due to the different saturation voltages of the L298's outputs, I do not believe that the load current will be equally divided between the two bridges; it's quite likely that one bridge output will carry more than 50% of the load. If anyone has an argument (other than the datasheet statement) why paralleling ought to be inherently safe, please educate me. Thanks.
 

ronsimpson

Joined Oct 7, 2019
1,020
please educate me
I am probably not qualified to educate you ...… but
Because he is paralleling to drivers in the same IC, the drivers are at the same temperature and the drivers were made at the same moment in time and the IC is being turn on/off at a slow rate, and (some other good argument I can not think of at this time), it is probably OK.
I agree that paralleling silicon is generally not a good idea.
 

TeeKay6

Joined Apr 20, 2019
572
I am probably not qualified to educate you ...… but
Because he is paralleling to drivers in the same IC, the drivers are at the same temperature and the drivers were made at the same moment in time and the IC is being turn on/off at a slow rate, and (some other good argument I can not think of at this time), it is probably OK.
I agree that paralleling silicon is generally not a good idea.
@ronsimpson
Thanks, Ron. Your comments are, I believe, true. Unfortunately the datasheet gives us no idea how close the matches between the two channels might, or might not, be. For most silicon IC's, expecting matching where none is specified is risky.

Consider a simple dual opamp with a max input offset of +/-10mV (for each channel). Presumably the matched input transistors for each channel are located very close together, at the same temperature, made at the same time...yet they exhibit mismatch to such a degree as to very often be the limiting factor for the opamp. Also note, the separation of the input transistors of one channel from those of the other channel is only a bit greater than between the two transistors of either single channel. Surely we could expect better matching between the two separate channels than between the two inputs of one channel? Of course not. To expect the two separate channels to have better matching than a single channel would be silly. That is, if we find that the input offset of one channel is +2mV, we are in no way entitled to assume the other channel will also have an input offset of about +2mV. However, this is a similar situation to what exists within the L298...except that the chip size (and consequent separations) of the L298 likely dwarfs the size of a simple dual opamp chip.
 

Thread Starter

rodolfoviolac

Joined Dec 3, 2019
4
Hi guys,

Thanks for the support, I made some changes and fully worked, I did a lot of tests and it worked perfectly.

Changes that I made:
- I Changed the diodes to Schottky switching diodes (1N5819 1A 40V).
- I Added two Rs Resistors (0.56 ohm 1W). One on SENSE_A and the other one on SENSE_B.
- I Changed the heat sink that I had to a large one.

The Rs resistors is very importante step, because they will help divide the current more equally to the both bridge that are in parallel, just to make sure that one bridge is not receiving more current that can really handle. I think all the changes above made the project work.

I'm very thankful for you all support!
 
Top