Date
1 - 20 of 34
1502/1503 Charging Discrepancy
|
Dave Casey
I was dusting off my TDR and charging up the battery pack I rebuilt a few
years back when I noticed something interesting. In the 1982 revisions of both the 1502 and 1503 manuals, it clearly states that the battery will not be overcharged if the unit is left plugged in past the maximum 16 hour recharge time. In the 1989 revision of the 1502 manual, it states exactly the opposite. The A6 power supply board didn't change between those revisions. The Theory of Operation sections regarding the battery charger have changed between the different revisions, but only in phrasing. The content remains the same, offering no explanation as to what does or does not overcharge the battery. Presumably the later information is the better information, but does anyone know the story behind why it changed? Dave Casey
|
|
|
|
Tom Gardner
My speculation would be that
toggle quoted messageShow quoted text
- the mean charging current might be C/10, and is therefore a suitable trickle charge rate - but the peak current in the charging pulses is much higher, and that somehow damages the cells If that's the case, I don't see any good way of avoiding overcharging the battery, unless you know the battery's charge state before you start charging. You can change the mean charging current by changing R6131, since the action of the PSU is to keep the average voltage across R6131 constant. While I really like my 1502, I hate that part of the PSU!
On 27/09/19 05:47, Dave Casey wrote:
I was dusting off my TDR and charging up the battery pack I rebuilt a few
|
|
|
|
Jürgen Florenkowski
Hello Dave,
when I bought my used 1502c years ago, a lead gel accumulator had been installed. This was broken, because according to the sign on the battery this Type is not suitable for long-term use. I installed a replacement, which is also installed in uninterruptible power supplies. This accumulator can be charged permanently with constant current. For these accumulators, the continuous charging current is specified in the data sheet. The current can then be adjusted in the power supply of the 1502c if necessary by a resistance change on the LM317 regulator. The value can be found in the data sheet of the LM317 (or similar IC). Maybe Tektronix had switched the accumulator type to a continuous charge type and therefore changed the manual. Greetings Juergen Sorry for my poor englich...
|
|
|
|
On Fri, Sep 27, 2019 at 06:47 AM, Dave Casey wrote:
Presumably the later information is the better information, but does anyonehttp://www.hakanh.com/dl/docs/hardtofind/1502batt.pdf /Håkan
|
|
|
|
Craig Sawyers <c.sawyers@...>
http://www.hakanh.com/dl/docs/hardtofind/1502batt.pdfYour ability to pull these gems out of your archives never fails to amaze me, Hakan! Kind regards Craig
|
|
|
|
Tom Gardner
"our engineers have researched the 1500 charging circuits and have determined..."
toggle quoted messageShow quoted text
is confirmation rather than a reason why. Maybe they were embarrassed, or afraid of repair costs and/or lawsuits. Pushing the problem onto the user is often the cheapest option :(
On 27/09/19 09:54, zenith5106 wrote:
On Fri, Sep 27, 2019 at 06:47 AM, Dave Casey wrote:Presumably the later information is the better information, but does anyonehttp://www.hakanh.com/dl/docs/hardtofind/1502batt.pdf
|
|
|
|
Chuck Harris <cfharris@...>
Good golly! That TELEX from tektronix is just about
toggle quoted messageShow quoted text
as clear as mud. Allow me to clarify: All NiCd's when charged at C/10 and higher will be damaged when left on to charge continuously. It is just a matter of when they will cease to be useful. If you want to do the best job possible of protecting your large C/10 charged packs, use a timed outlet. You can usually find them right alongside of the interval timer outlets, used to turn a house lamp on and off on a cycle. -Chuck Harris zenith5106 wrote:
On Fri, Sep 27, 2019 at 06:47 AM, Dave Casey wrote:Presumably the later information is the better information, but does anyonehttp://www.hakanh.com/dl/docs/hardtofind/1502batt.pdf
|
|
|
|
So, I have to ask: What charge rate is OK for long term float of Ni-Cd cells?
toggle quoted messageShow quoted text
D.
-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Chuck Harris Sent: 27 September 2019 12:34 To: TekScopes@groups.io Subject: Re: [TekScopes] 1502/1503 Charging Discrepancy Good golly! That TELEX from tektronix is just about as clear as mud. Allow me to clarify: All NiCd's when charged at C/10 and higher will be damaged when left on to charge continuously. It is just a matter of when they will cease to be useful. If you want to do the best job possible of protecting your large C/10 charged packs, use a timed outlet. You can usually find them right alongside of the interval timer outlets, used to turn a house lamp on and off on a cycle. -Chuck Harris zenith5106 wrote: On Fri, Sep 27, 2019 at 06:47 AM, Dave Casey wrote:Presumably the later information is the better information, but does anyonehttp://www.hakanh.com/dl/docs/hardtofind/1502batt.pdf
|
|
|
|
Tom Gardner
From memory, I measured the mean charging current to be about C/10, i.e. 140mA with ~1.6Ah cells. There's some uncertainty about that due to the highly pulsed nature of the charging waveform.
toggle quoted messageShow quoted text
The timer solution fails if you don't know the state of charge when starting to charge. Obviously if it is initially, say, 3/4 charged then a full 16 hours charge would be inappropriate! Calculating the initial state of charge wasn't (and isn't) easy.
On 27/09/19 12:34, Chuck Harris wrote:
Good golly! That TELEX from tektronix is just about
|
|
|
|
Tom Gardner
That's an RTFDS question, but significantly complicated by 1502/3 not having a constant charging current.
toggle quoted messageShow quoted text
A constant non-pulsed C/10 current is the traditional rate stated for NiCds at room temperature.
On 27/09/19 13:07, David C. Partridge wrote:
So, I have to ask: What charge rate is OK for long term float of Ni-Cd cells?
|
|
|
|
Harvey White
IIRC, it's C/20 or less.
toggle quoted messageShow quoted text
A smart charger can charge at C, C/2, C/4 until either voltage or temperature rise, then shift to C/20 (?) for top off, then shift to less than that for maintenance float. I'd suggest the BQ2000 series or the BQ24000 series for interesting reading. One is switching, the 2000 series is linear. Charge rates and the like are adjustable to some extents. A bit of delving on various auction sites may find a pre-built board that solves the problem in a small package. Harvey
On 9/27/2019 8:07 AM, David C. Partridge wrote:
So, I have to ask: What charge rate is OK for long term float of Ni-Cd cells?
|
|
|
|
Tom Gardner
I've seen C/20 for NiMH cells at room temperature, not for NiCd.
toggle quoted messageShow quoted text
Of course the 1502/3 charger is the antithesis of "intelligent"!
On 27/09/19 14:25, Harvey White wrote:
IIRC, it's C/20 or less.
|
|
|
|
Chuck Harris <cfharris@...>
None. However, all NiCd's do self discharge, so one could
toggle quoted messageShow quoted text
infer from that that the self discharge rate would be safe. All engineering is a balance of perfect vs good enough, salted with: Can we afford it? Ideally, no battery should be float or trickle charged if longest life is the only criterion. Longest life usually comes from bursts of charging separated by a significant length of time. For NiCD's, something like a full trickle charge every 3 months works quite well. I did a complicated charger like that for one project, where there were multiple states. The float state trickle charged every other month, and there was a power loss state that timed the power loss, and quick charged based on the length of time the power was off. Sounds complicated, but all of the control stuff fit in one 8 pin PIC processor ($0.99), and took a couple of pages of "C" code to describe the operation. Charge rate was handled by a LM317 configured as a constant current source with two rates: trickle, and quick charge. Whatever you do, don't attempt to discharge a NiCd battery after it has sat about, without first doing a full trickle charge. Each cell will self discharge at its own rate, and a long time sitting will result in a battery string that has every cell with different amounts of charge. If you attempt to use the device, it is inevitable that one cell will run out of charge first, and the other cells will reverse charge that weak cell, drastically shortening life expectancy. Chuck Harris David C. Partridge wrote:
So, I have to ask: What charge rate is OK for long term float of Ni-Cd cells?
|
|
|
|
Chuck Harris <cfharris@...>
RTFDS, I haven't heard that in a while...
toggle quoted messageShow quoted text
NiCds really, really, like to be pulse charged. The most effective known method of quick charging a NiCd pack is to use Christie's (now expired) patented burp charge method, called a reflex charge. The burp charge provides a very high current pulse for about 100ms, followed by a discharge pulse for about 5ms, repeated every second. Length of charge is controlled by measuring the cell voltage during the rest time between pulses, and looking for a change to negative slope in the cell voltage... commonly known as -dV charging. A refinement can be had by instead taking the second derivative and finding the inflection point, where the slope of the cell voltage starts to change, and quitting at that point. The high current charge gets the current into the cell quickly, but releases little bubbles of steam from the plates. The momentary discharge pulse causes the bubbles to recombine back to water in the electrolyte, and life goes on. -Chuck Harris Tom Gardner wrote:
That's an RTFDS question, but significantly complicated by 1502/3 not having a
|
|
|
|
Tom Gardner
The 1502/1503 PSU protects the battery by turning off the SMPS when the battery voltage is too low. That should be sufficient to prevent cells being reverse-charged.
toggle quoted messageShow quoted text
Unfortunately on/off switch doesn't disconnect the chart recorder. Hence, even when switched off, the battery is discharged at 1-2mA. That can result in cells being reverse charged. To counter that, the manual states that the battery should be recharged every month.
On 27/09/19 14:38, Chuck Harris wrote:
None. However, all NiCd's do self discharge, so one could
|
|
|
|
Tom Gardner
Interesting, but it doesn't help in the case of the 1502/3 charger, which simply senses and controls the mean charging current, fixed at ~C/10.
toggle quoted messageShow quoted text
On 27/09/19 14:54, Chuck Harris wrote:
RTFDS, I haven't heard that in a while...
|
|
|
|
Harvey White
if I used mine more, I'd replace the charger with something smart.
toggle quoted messageShow quoted text
I'd still be tempted to do that on my 214, if I could ever find a good vertical board (switch is messed up). In fact, given how the 200 series works, I'd almost insist on doing it. Those chips were not available when the equipment was designed, I think, and the knowledge of exactly how batteries work and why wasn't as developed. Any current not used to reverse the chemistry only heats the cell, which doesn't help. That's one reason why the delta T algorithm works on Nicad/Nimh cells. There's also a fractional voltage rise that can be sensed as well. Harvey
On 9/27/2019 9:34 AM, Tom Gardner wrote:
I've seen C/20 for NiMH cells at room temperature, not for NiCd.
|
|
|
|
Chuck Harris <cfharris@...>
Sadly it doesn't.
toggle quoted messageShow quoted text
A fully charged NiCd cell is slightly more than 1.4V, but drops down to 1.25V for the bulk of its charge life. The battery protection circuitry in the 1502/3 TDR is set for battery shutdown at 9.5V. There are 9 cells in the 1502/3 battery, so, if all 9 cells are equal, shutdown occurs at 1.06V/cell... which sounds good... in principle... However, let's assume one cell is weak, and fully discharges before the others: 9.5Vcutoff / 8 cells = 1.19Vcutoff/cell If one weak cell hits 0V at any point before the strong cells are below 1.19V each, the strong cells will reverse charge that one weak cell until they collectively drop below 1.19V/cell. Given an extremely likely situation where the weak cell hits zero while the strong cells are still well above their nominal 1.25V cell voltage, the reversed cell has to reach -0.5V before the cutoff will be tripped. A 9 cell pack is a whole lot safer than an 18 cell pack, but absent perfectly matched cells, it can still ruin the weakest cell before the battery protection circuit cuts out with 1.06V/cell cutoff (9.5V). Adding a Schottky diode across each cell to protect against reverse charge may reduce the damage, but won't eliminate it. -Chuck Harris Tom Gardner wrote:
The 1502/1503 PSU protects the battery by turning off the SMPS when the battery
|
|
|
|
Tom Gardner
The manual states the 1502 cutoff battery voltage is 10V, not 9.5V. ISTR that being about right on my 1502. That ameliorates your statement to some extent;
toggle quoted messageShow quoted text
The degree of imbalance will always be an issue, and the result will always be imperfect. Nonetheless the "protection" is better than nothing, and better than most equipment of that era (and for a couple of decades afterwards!). The chart recorder constant drain is a real problem; I remove my battery pack when not in use.
On 27/09/19 16:41, Chuck Harris wrote:
Sadly it doesn't.
|
|
|
|
Chuck Harris <cfharris@...>
The Christie papers talk of pulse charging reducing the
toggle quoted messageShow quoted text
regularity of crystalline structures formed during charging. They claim that the regular structure reduces the cell capacity over time. Pulse charging restores the cell to its original cell capacity. Naturally, they are reporting their experiences with the reflex pulse charging scheme they developed from their NiCd research. I don't think the 1502's 60Hz pulsing bothers the cells one bit... assuming that you make sure that the RMS charging current is always less than C/10, and don't simply leave the TDR plugged in always. That said, there has to be some threshold point where the pulse current can be so extreme that even though the RMS value is reasonable, damage can occur... I don't think the 1502/3 is anywhere near that level. C/10 is a number thrown out by the cell manufacturers as a good balance point between user convenience, and ultimate cell life. At C/10, the pack will survive most of the abuse of limited overcharge, while still maintaining a fairly cost effective cell life. They are counting on the user not to simply leave the pack charging for months on end... but rather to actively charge it, and use it. -Chuck Harris Tom Gardner wrote:
Interesting, but it doesn't help in the case of the 1502/3 charger, which simply
|
|
|