View Single Post
Old 14th Aug 2006, 9:26 pm   #1
YC-156
Retired Dormant Member
 
YC-156's Avatar
 
Join Date: May 2005
Location: Aarhus, Denmark
Posts: 281
Default Building Block: 50MHz prescaler for LF frequency counter.

As mentioned elsewhere, then this weekend past I made a little prescaler for the frequency counter in my new DMM. This is a rather brutal job in the simplicity department, bordering on being crude, yet it works quite well after a bit of tweaking of the buffer amplifier stages.

There are not many surprises in the circuit, except perhaps for the 7805 making a showing in a battery powered piece of equipment. Not having an 78L05, the 100mA low power version of the 7805, in my junkbox, I tested the current consumption of a random 7805 and found it to be only 3.2mA. That is less than the guaranteed maximum of an 78L05, so in went the 7805.

If I had another 'HCT, then I would have used it in place of the ordinary CD4017, since the HCT has a better drive capability of the counter following the prescaler. Even at a few hundred KHz the output is running at, it is possible to see that the corners of the square wave are rounded, probably due to a combination of insufficient drive from the 4017 and stray capacitances.

The current consumed varies between 7.0 mA at idle (no input signal) and 17.0mA when fed 50MHz from a signal generator. That should allow for around 30 hours of continuous operation on the single 9V battery shown in the photo, worst case.

By doing the math it is easy to see that unsurprisingly the 7 mA of idle current is almost completely consumed by the 7805 and the two buffer stages, with next to nothing left for the two ICs.

Initially I had pressed a BF199 into service where the BFR91 now resides. That wasn't a wise desicion though. I had forgotten about the transition frequency of transistors, ft, falling as the current is lowered. Tweak as I might the circuit simply ran out of oomph at around 30MHz.

The BFR91 is a different story, of course, and it apparently doesn't mind chopping DC at 55-60MHz, even though I lowered the current to half that used for the BF199. At around 60MHz the operation of the counter begins to become unstable, either due to low signal amplitude (again), or simply because I hit the frequency limit of the 'HCT. Its typical maximum frequency of operation is 67MHz. The highest reliable count I have squeezed out of it so far is 58MHz.

The low collector voltage of the BFR91 at idle is no mistake. I tweaked the stage using the scope for maximum voltage swing and getting as close as possible to a 50% duty cycle on the output waveform. That yielded the resistor values shown in the diagram.

Bonus exercise: Calculate the DC current gain, hfe, for this particular BFR91 specimen. It is guaranteed to be at least 40 (at Ic = 30mA).

Frank N.
Attached Thumbnails
Click image for larger version

Name:	prescaler_cir.jpg
Views:	480
Size:	112.9 KB
ID:	4961  
YC-156 is offline