Maker Pro
Maker Pro

Bike Horn

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
BikeHornPCB1.JPG Ok, so I buy two Chinese bike horns on ebay. One has a brand name of "RockBros" and another that's generic but identical to the RockBros model.

Originally, I was going to start this topic in our uC section for those that may be interested in doing what I plan to do. That being,.. replace the uC with my own that contains my own program. This is because I only like 1 of the 5 programmed sounds that they provided.

Fast forward: I decided that I needed to reverse engineer the circuit so I could can replicate the PIC pin-out and program my PIC accordingly. I've attached my reverse engineered schematic and photos of the board, so give them a peek.

Take note of the Piezo driver circuit Q1 and T1. T1 is a step-up auto-transformer that increases the audio output voltage to the Piezo. Nothing really odd there except for the fact that the audio signal from the uC will be a digital signal with fast rising and falling edges. This is what's got me scratching my head, as I don't see anything in this circuit to protect Q1 (or any other component) from back EMF that T1 will certainly produce.

Admittedly, I have not scoped this circuit as yet. Hopefully, doing so may bring some enlightenment,.. like a protection diode integral to T1. Though, I would think that a reversed biased diode across Q1's collector - emitter junction would be where I'd put it.

Any thoughts?

ChrisBikeHornSch.JPG BikeHornPCB2.jpg
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,682
Joined
Nov 17, 2011
Messages
13,682
I don't see anything in this circuit to protect Q1 (or any other component) from back EMF that T1 will certainly produce.
Back EMF is probably not a big issue here as the main part of it will develop between pins 1 and 2 of the transformer (at least if the winding ratios are as shown in your schematic).
On the other hand the Chinese manufacturer may not have considered back EMF at all.
Anyway adding a freewheeling diode shouldn't be an issue, will it?
 

Colin Mitchell

Aug 31, 2014
1,416
Joined
Aug 31, 2014
Messages
1,416
There are no spikes passed to the transistor. Look at the circuit.
The piezo needs about 80v.
 

poor mystic

Apr 8, 2011
1,074
Joined
Apr 8, 2011
Messages
1,074
Possibly this is one of your device's "intentional failure modes".
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
Back EMF is probably not a big issue here as the main part of it will develop between pins 1 and 2 of the transformer (at least if the winding ratios are as shown in your schematic).
On the other hand the Chinese manufacturer may not have considered back EMF at all.
Anyway adding a freewheeling diode shouldn't be an issue, will it?

I'm not sure of a number of things. The Xfmr may not be wired exactly as I've drawn it so I may be revising the schematic.

In the past I've made it a habit of giving a big "UGH!" to anything posted from the Instructables website. But like the old saying ... "Sometimes, even a blind Squirrel will find the nut." a miracle happened while doing a Google search. The transformer in this link looks exactly like mine but that doesn't mean it is. Odds are that it is though.
http://www.instructables.com/id/How-to-make-an-Arduino-driven-Piezo-LOUD/

The link also contains other links which brought up more (very similar) transformer data.
http://www.ozitronics.com/docs/k15.pdf

One thing seems common in these links is that there's no flyback diode but spicing the driver circuit shows spikes that would certainly kill Q1 in a heartbeat. Admittedly, the load I'm using to represent the Piezo is just a 1K resistor, which doesn't represent a piezo. Is the capacitance shown in piezo datasheets representative of the series or parallel capacitance?

Thanks,
Chris
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
There are no spikes passed to the transistor. Look at the circuit.
The piezo needs about 80v.

While reading through Piezo datasheets I haven't found any that spec anything over a 20V-PP square wave. This link indicates 200VPP but I believe it's a decimal place typo.
http://www.discovercircuits.com/DJ-Circuits/beeper4.htm

As for my concern about back EMF on the primary .. It's very possible that the relatively large parallel capacitance of a piezo may swamp out the back EMF on the primary.

Chris
 
Last edited:
Top